Scholars@Duke
Vahid Tarokh

Vahid Tarokh

Rhodes Family Distinguished Professor of Electrical and Computer Engineering
Pierre R. Lamond Department of Electrical and Computer Engineering
vahid.tarokh@duke.edu
Box 90291, Durham, NC 27708
130 Hudson Hall, Durham, NC 27708

Scholarly Works - Conferences

ELLIPTIC LOSS REGULARIZATION

Conference 13th International Conference on Learning Representations Iclr 2025 · January 1, 2025 Regularizing neural networks is important for anticipating model behavior in regions of the data space that are not well represented. In this work, we propose a regularization technique for enforcing a level of smoothness in the mapping between the data in ... Cite

Variational Adversarial Training Towards Policies with Improved Robustness

Conference Proceedings of Machine Learning Research · January 1, 2025 Reinforcement learning (RL), while being the benchmark for policy formulation, often struggles to deliver robust solutions across varying scenarios, leading to marked performance drops under environmental perturbations. Traditional adversarial training, ba ... Cite

Parabolic Continual Learning

Conference Proceedings of Machine Learning Research · January 1, 2025 Regularizing continual learning techniques is important for anticipating algorithmic behavior under new realizations of data. We introduce a new approach to continual learning by imposing the properties of a parabolic partial differential equation (PDE) to ... Cite

Steinmetz Neural Networks for Complex-Valued Data

Conference Proceedings of Machine Learning Research · January 1, 2025 We introduce a new approach to processing complex-valued data using DNNs consisting of parallel real-valued subnetworks with coupled outputs. Our proposed class of architectures, referred to as Steinmetz Neural Networks, incorporates multi-view learning to ... Cite

CATE Estimation With Potential Outcome Imputation From Local Regression

Conference Proceedings of Machine Learning Research · January 1, 2025 One of the most significant challenges in Conditional Average Treatment Effect (CATE) estimation is the statistical discrepancy between distinct treatment groups. To address this issue, we propose a model-agnostic data augmentation method for CATE estimati ... Cite

Conditional Average Treatment Effect Estimation Under Hidden Confounders

Conference Proceedings of Machine Learning Research · January 1, 2025 One of the major challenges in estimating conditional potential outcomes and conditional average treatment effects (CATE) is the presence of hidden confounders. Since testing for hidden confounders cannot be accomplished only with observational data, condi ... Cite

In-Context Reinforcement Learning From Suboptimal Historical Data

Conference Proceedings of Machine Learning Research · January 1, 2025 Transformer models have achieved remarkableempirical successes, largely due to their incontext learning capabilities. Inspired by this, we explore training an autoregressive transformer for in-context reinforcement learning (ICRL). In this setting, we init ... Cite

Decoding Rewards in Competitive Games: Inverse Game Theory with Entropy Regularization

Conference Proceedings of Machine Learning Research · January 1, 2025 Estimating the unknown reward functions driving agents’ behavior is a central challenge in inverse games and reinforcement learning. This paper introduces a unified framework for reward function recovery in two-player zero-sum matrix games and Markov games ... Cite

Pi-talk: Edge-Only, Adapter-Tuned Multimodal Small Language Model for Safe, Real-Time In-Vehicle Dialogue

Conference Proceedings 2025 24th International Conference on Machine Learning and Applications Icmla 2025 · January 1, 2025 Natural-language interaction between passengers and autonomous vehicles is essential for trust, safety, and user experience, but deploying Large Language Models (LLMs) on automotive edge platforms is constrained by compute, memory, energy, and privacy. We ... Full text Cite

Neural McKean-Vlasov Processes: Distributional Dependence in Diffusion Processes

Conference Proceedings of Machine Learning Research · January 1, 2024 McKean-Vlasov stochastic differential equations (MV-SDEs) provide a mathematical description of the behavior of an infinite number of interacting particles by imposing a dependence on the particle density. We study the influence of explicitly including dis ... Cite

REFORMA: Robust REinFORceMent Learning via Adaptive Adversary for Drones Flying under Disturbances

Conference Proceedings IEEE International Conference on Robotics and Automation · January 1, 2024 In this work, we introduce REFORMA, a novel robust reinforcement learning (RL) approach to design controllers for unmanned aerial vehicles (UAVs) robust to unknown disturbances during flights. These disturbances, typically due to wind turbulence, electroma ... Full text Cite

Base Models for Parabolic Partial Differential Equations

Conference Proceedings of Machine Learning Research · January 1, 2024 Parabolic partial differential equations (PDEs) appear in many disciplines to model the evolution of various mathematical objects, such as probability flows, value functions in control theory, and derivative prices in finance. It is often necessary to comp ... Cite

Distributionally Robust Optimization as a Scalable Framework to Characterize Extreme Value Distributions

Conference Proceedings of Machine Learning Research · January 1, 2024 The goal of this paper is to develop distributionally robust optimization (DRO) estimators, specifically for multidimensional Extreme Value Theory (EVT) statistics. EVT supports using semi-parametric models called max-stable distributions built from spatia ... Cite

Random Linear Projections Loss for Hyperplane-Based Optimization in Neural Networks

Conference Proceedings of Machine Learning Research · January 1, 2024 Advancing loss function design is pivotal for optimizing neural network training and performance. This work introduces Random Linear Projections (RLP) loss, a novel approach that enhances training efficiency by leveraging geometric relationships within the ... Cite

Steering Decision Transformers via Temporal Difference Learning

Conference IEEE International Conference on Intelligent Robots and Systems · January 1, 2024 Decision Transformers (DTs) have been highly effective for offline reinforcement learning (RL) tasks, successfully modeling the sequences of actions in a given set of demonstrations. However, DTs may perform poorly in stochastic environments, which are pre ... Full text Cite

DynamicFL: Federated Learning with Dynamic Communication Resource Allocation

Conference Proceedings 2024 IEEE International Conference on Big Data Bigdata 2024 · January 1, 2024 Federated Learning (FL) is a collaborative machine learning framework that allows multiple users to train models utilizing their local data in a distributed manner. However, considerable statistical heterogeneity in local data across devices often leads to ... Full text Cite

PERCEIVING COPULAS FOR MULTIMODAL TIME SERIES FORECASTING

Conference Proceedings Winter Simulation Conference · January 1, 2024 Transformers have demonstrated remarkable efficacy in forecasting time series. However, their dependence on self-attention mechanisms demands significant computational resources, thereby limiting their applicability across diverse tasks. Here, we propose t ... Full text Cite

GENERATIVE LEARNING FOR SIMULATION OF VEHICLE FAULTS

Conference Proceedings Winter Simulation Conference · January 1, 2024 We develop a novel generative model to simulate vehicle health and forecast faults, conditioned on practical operational considerations. The model, trained on data from the US Army’s Predictive Logistics program, aims to support predictive maintenance. It ... Full text Cite

Multimodal Controller for Generative Models

Conference Lecture Notes in Networks and Systems · January 1, 2023 Class-conditional generative models are crucial tools for data generation from user-specified class labels. Existing approaches for class-conditional generative models require nontrivial modifications of backbone generative architectures to model condition ... Full text Cite

Subspace Perturbation Analysis for Data-Driven Radar Target Localization

Conference Proceedings of the IEEE Radar Conference · January 1, 2023 Recent works exploring data-driven approaches to classical problems in adaptive radar have demonstrated promising results pertaining to the task of radar target localization. Via the use of space-time adaptive processing (STAP) techniques and convolutional ... Full text Cite

Score-based Quickest Change Detection for Unnormalized Models

Conference Proceedings of Machine Learning Research · January 1, 2023 Classical change detection algorithms typically require modeling pre-change and post-change distributions. The calculations may not be feasible for various machine learning models because of the complexity of computing the partition functions and normalize ... Cite

Robust Quickest Change Detection for Unnormalized Models

Conference Proceedings of Machine Learning Research · January 1, 2023 Detecting an abrupt and persistent change in the underlying distribution of online data streams is an important problem in many applications. This paper proposes a new robust score-based algorithm called RSCUSUM, which can be applied to unnormalized models ... Cite

Inference and Sampling of Point Processes from Diffusion Excursions

Conference Proceedings of Machine Learning Research · January 1, 2023 Point processes often have a natural interpretation with respect to a continuous process. We propose a point process construction that describes arrival time observations in terms of the state of a latent diffusion process. In this framework, we relate the ... Cite

Transfer Learning for Individual Treatment Effect Estimation

Conference Proceedings of Machine Learning Research · January 1, 2023 This work considers the problem of transferring causal knowledge between tasks for Individual Treatment Effect (ITE) estimation. To this end, we theoretically assess the feasibility of transferring ITE knowledge and present a practical framework for effici ... Cite

PASTA: Pessimistic Assortment Optimization

Conference Proceedings of Machine Learning Research · January 1, 2023 We consider a class of assortment optimization problems in an offline data-driven setting. A firm does not know the underlying customer choice model but has access to an offline dataset consisting of the historically offered assortment set, customer choice ... Cite

Off-Policy Evaluation for Human Feedback

Conference Advances in Neural Information Processing Systems · January 1, 2023 Off-policy evaluation (OPE) is important for closing the gap between offline training and evaluation of reinforcement learning (RL), by estimating performance and/or rank of target (evaluation) policies using offline trajectories only. It can improve the s ... Cite

PRUNING DEEP NEURAL NETWORKS FROM A SPARSITY PERSPECTIVE

Conference 11th International Conference on Learning Representations Iclr 2023 · January 1, 2023 In recent years, deep network pruning has attracted significant attention in order to enable the rapid deployment of AI into small devices with computation and memory constraints. Pruning is often achieved by dropping redundant weights, neurons, or layers ... Cite

CHARACTERISTIC NEURAL ORDINARY DIFFERENTIAL EQUATIONS

Conference 11th International Conference on Learning Representations Iclr 2023 · January 1, 2023 We propose Characteristic-Neural Ordinary Differential Equations (C-NODEs), a framework for extending Neural Ordinary Differential Equations (NODEs) beyond ODEs. While NODE models the evolution of latent variables as the solution to an ODE, C-NODE models t ... Cite

Online Action Change Detection for Automatic Vision-based Ground Control of Aircraft

Conference AIAA Science and Technology Forum and Exposition AIAA Scitech Forum 2022 · January 1, 2022 Classical action recognition algorithms require the user to pre-select the time window for the action by clipping the video or choosing the initial and final time frames. Recently, new deep learning algorithms have been developed to detect a key action fro ... Full text Cite

Toward Data-Driven STAP Radar

Conference Proceedings of the IEEE Radar Conference · January 1, 2022 Using an amalgamation of techniques from classical radar, computer vision, and deep learning, we characterize our ongoing data-driven approach to space-time adaptive processing (STAP) radar. We generate a rich example dataset of received radar signals by r ... Full text Cite

A Physics-Informed Vector Quantized Autoencoder for Data Compression of Turbulent Flow

Conference Data Compression Conference Proceedings · January 1, 2022 Analyzing large-scale data from simulations of turbulent flows is memory intensive, requiring significant resources. This major challenge highlights the need for data compression techniques. In this study, we apply a physics-informed Deep Learning techniqu ... Full text Cite

On The Energy Statistics of Feature Maps in Pruning of Neural Networks with Skip-Connections

Conference Data Compression Conference Proceedings · January 1, 2022 We propose a new structured pruning framework for compressing Deep Neural Networks (DNNs) with skip-connections, based on measuring the statistical dependency of hidden layers and predicted outputs. The dependence measure defined by the energy statistics o ... Full text Cite

Multi-Agent Adversarial Attacks for Multi-Channel Communications

Conference Proceedings of the International Joint Conference on Autonomous Agents and Multiagent Systems Aamas · January 1, 2022 Recently Reinforcement Learning (RL) has been applied as an anti-adversarial remedy in wireless communication networks. However studying the RL-based approaches from the adversary's perspective has received little attention. Additionally, RL-based approach ... Cite

Towards Deep Learning-Guided Multiuser SNR and Doppler Shift Detection for Next-Generation Wireless Systems

Conference IEEE Vehicular Technology Conference · January 1, 2022 In order to meet the ever-growing demand for data traffic, highly efficient multiple access schemes, such as OFDMA, are widely used in modern communication standards. In such multiple access schemes, adaptive modulation and coding (AMC) are used to optimiz ... Full text Cite

Semi-Empirical Objective Functions for MCMC Proposal Optimization

Conference Proceedings International Conference on Pattern Recognition · January 1, 2022 Current objective functions used for training neural MCMC proposal distributions implicitly rely on architectural restrictions to yield sensible optimization results, which hampers the development of highly expressive neural MCMC proposal architectures. In ... Full text Cite

Modeling Extremes with d-max-decreasing Neural Networks

Conference Proceedings of the 38th Conference on Uncertainty in Artificial Intelligence Uai 2022 · January 1, 2022 We propose a neural network architecture that enables non-parametric calibration and generation of multivariate extreme value distributions (MEVs). MEVs arise from Extreme Value Theory (EVT) as the necessary class of models when extrapolating a distributio ... Cite

Personalized Federated Recommender Systems with Private and Partially Federated AutoEncoders

Conference Conference Record Asilomar Conference on Signals Systems and Computers · January 1, 2022 Recommender Systems (RSs) have become increasingly important in many application domains, such as digital marketing. Conventional RSs often need to collect users' data, centralize them on the server-side, and form a global model to generate reliable recomm ... Full text Cite

BLASCHKE PRODUCT NEURAL NETWORK (BPNN): A PHYSICS-INFUSED NEURAL NETWORK FOR PHASE RETRIEVAL OF MEROMORPHIC FUNCTIONS

Conference Iclr 2022 10th International Conference on Learning Representations · January 1, 2022 Numerous physical systems are described by ordinary or partial differential equations whose solutions are given by holomorphic or meromorphic functions in the complex domain. In many cases, only the magnitude of these functions are observed on various poin ... Cite

TASK AFFINITY WITH MAXIMUM BIPARTITE MATCHING IN FEW-SHOT LEARNING

Conference Iclr 2022 10th International Conference on Learning Representations · January 1, 2022 We propose an asymmetric affinity score for representing the complexity of utilizing the knowledge of one task for learning another one. Our method is based on the maximum bipartite matching algorithm and utilizes the Fisher Information matrix. We provide ... Cite

Modeling Extremes with d-max-decreasing Neural Networks

Conference Proceedings of Machine Learning Research · January 1, 2022 We propose a neural network architecture that enables non-parametric calibration and generation of multivariate extreme value distributions (MEVs). MEVs arise from Extreme Value Theory (EVT) as the necessary class of models when extrapolating a distributio ... Cite

Inference and Sampling for Archimax Copulas

Conference Advances in Neural Information Processing Systems · January 1, 2022 Understanding multivariate dependencies in both the bulk and the tails of a distribution is an important problem for many applications, such as ensuring algorithms are robust to observations that are infrequent but have devastating effects. Archimax copula ... Cite

SemiFL: Semi-Supervised Federated Learning for Unlabeled Clients with Alternate Training

Conference Advances in Neural Information Processing Systems · January 1, 2022 Federated Learning allows the training of machine learning models by using the computation and private data resources of many distributed clients. Most existing results on Federated Learning (FL) assume the clients have ground-truth labels. However, in man ... Cite

GAL: Gradient Assisted Learning for Decentralized Multi-Organization Collaborations

Conference Advances in Neural Information Processing Systems · January 1, 2022 Collaborations among multiple organizations, such as financial institutions, medical centers, and retail markets in decentralized settings are crucial to providing improved service and performance. However, the underlying organizations may have little inte ... Cite

Knowledge-Aided Data-Driven Radar Clutter Representation

Conference IEEE National Radar Conference Proceedings · May 7, 2021 We use a knowledge-aided, data-driven, location-aware approach based on the RFView simulation software to model and estimate the effect of ground clutter in airborne radars. Using the RFView simulator, we produce many samples of potential clutter effects a ... Full text Cite

Compressing Deep Networks Using Fisher Score of Feature Maps

Conference Data Compression Conference Proceedings · March 1, 2021 In this paper, we propose a new structural technique for pruning deep neural networks with skip-connections. Our approach is based on measuring the importance of feature maps in predicting the output of the model using their Fisher scores. These scores sub ... Full text Cite

Flow parameter estimation based on on-board measurements of air vehicle traversing turbulent flows

Conference AIAA Scitech 2021 Forum · January 1, 2021 Inspired by principles from particle transport theory in fluid dynamics, we recently developed a new energy-efficient control approach via implicit model following (IMF) for air vehicles traversing turbulent flows. However, the control design requires prio ... Cite

Fisher Auto-Encoders

Conference Proceedings of Machine Learning Research · January 1, 2021 It has been conjectured that the Fisher divergence is more robust to model uncertainty than the conventional Kullback-Leibler (KL) divergence. This motivates the design of a new class of robust generative auto-encoders (AE) referred to as Fisher auto-encod ... Cite

HETEROFL: COMPUTATION AND COMMUNICATION EFFICIENT FEDERATED LEARNING FOR HETEROGENEOUS CLIENTS

Conference Iclr 2021 9th International Conference on Learning Representations · January 1, 2021 Federated Learning (FL) is a method of training machine learning models on private data distributed over a large number of possibly heterogeneous clients such as mobile phones and IoT devices. In this work, we propose a new federated learning framework nam ... Cite

Investigation of Input Signal Representation to CNN for Improving SNR Classification Accuracy

Conference IEEE Vehicular Technology Conference · January 1, 2021 With the increase in demand for wireless data traffic, high-speed communication systems are required in many environment. Adaptive modulation and coding is an important technique to realize this, but it requires feedback of communication environment inform ... Full text Cite

Generative Archimedean Copulas

Conference 37th Conference on Uncertainty in Artificial Intelligence Uai 2021 · January 1, 2021 We propose a new generative modeling technique for learning multidimensional cumulative distribution functions (CDFs) in the form of copulas. Specifically, we consider certain classes of copulas known as Archimedean and hierarchical Archimedean copulas, po ... Cite

Generative Archimedean Copulas

Conference Proceedings of Machine Learning Research · January 1, 2021 We propose a new generative modeling technique for learning multidimensional cumulative distribution functions (CDFs) in the form of copulas. Specifically, we consider certain classes of copulas known as Archimedean and hierarchical Archimedean copulas, po ... Cite

Benchmarking Data-driven Surrogate Simulators for Artificial Electromagnetic Materials

Conference Advances in Neural Information Processing Systems · January 1, 2021 Artificial electromagnetic materials (AEMs), including metamaterials, derive their electromagnetic properties from geometry rather than chemistry. With the appropriate geometric design, AEMs have achieved exotic properties not realizable with conventional ... Cite

Deep James-Stein Neural Networks for Brain-Computer Interfaces

Conference ICASSP IEEE International Conference on Acoustics Speech and Signal Processing Proceedings · May 1, 2020 Nonparametric regression has proven to be successful in extracting features from limited data in neurological applications. However, due to data scarcity, most brain-computer interfaces still rely on linear classifiers. This work leverages the robustness o ... Full text Cite

Proximal Gradient Algorithm with Momentum and Flexible Parameter Restart for Nonconvex Optimization

Conference PROCEEDINGS OF THE TWENTY-NINTH INTERNATIONAL JOINT CONFERENCE ON ARTIFICIAL INTELLIGENCE · 2020 Link to item Cite

On the information of feature maps and pruning of deep neural networks

Conference Proceedings International Conference on Pattern Recognition · January 1, 2020 A technique for compressing deep neural models achieving competitive performance to state-of-the-art methods is proposed. The approach utilizes the mutual information between the feature maps and the output of the model in order to prune the redundant laye ... Full text Cite

Recurrent Neural Network-Assisted Adaptive Sampling for Approximate Computing

Conference Proceedings 2019 IEEE International Conference on Big Data Big Data 2019 · December 1, 2019 We propose an adaptive signal sampling approach that dynamically adjusts the sampling rate to approximate the local Nyquist rate of the signal. The proposed adaptive sampling approach consists of a recurrent neural network-based change detector that detect ... Full text Cite

Multi-Level Mean-Shift Clustering for Single-Channel Radio Frequency Signal Separation

Conference IEEE International Workshop on Machine Learning for Signal Processing Mlsp · October 1, 2019 Emerging wireless communication applications have led to a crowded radio frequency (RF) spectrum. Therefore, it is desired to develop signal separation techniques that can extract different RF signals from their mixtures. Existing signal separation approac ... Full text Cite

Gradient information for representation and modeling

Conference Advances in Neural Information Processing Systems · January 1, 2019 Motivated by Fisher divergence, in this paper we present a new set of information quantities which we refer to as gradient information. These measures serve as surrogates for classical information measures such as those based on logarithmic loss, Kullback- ... Cite

Restricted Recurrent Neural Networks.

Conference IEEE BigData · 2019 Cite

SGD converges to global minimum in deep learning via star-convex path

Conference 7th International Conference on Learning Representations Iclr 2019 · January 1, 2019 Stochastic gradient descent (SGD) has been found to be surprisingly effective in training a variety of deep neural networks. However, there is still a lack of understanding on how and why SGD can train these complex networks towards a global minimum. In th ... Cite

A Penalized Method for the Predictive Limit of Learning

Conference ICASSP IEEE International Conference on Acoustics Speech and Signal Processing Proceedings · September 10, 2018 Machine learning systems learn from and make predictions by building models from observed data. Because large models tend to overfit while small models tend to underfit for a given fixed dataset, a critical challenge is to select an appropriate model (e.g. ... Full text Cite

Supervised Learning Using Data-dependent Random Features with Application to Seizure Detection

Conference Proceedings of the IEEE Conference on Decision and Control · July 2, 2018 The randomized-feature technique has been successfully applied to large-scale supervised learning. Despite being significantly more efficient compared to kernel methods in terms of computational cost, random features can be improved from generalization (pr ... Full text Cite

Optimal prediction of data with unknown abrupt change points

Conference 2017 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2017 - Proceedings · March 7, 2018 We develop a novel methodology for predicting time series under unknown abrupt changes in data generating distributions. Based on Kolmogorov and Tikhomirov's e entropy, we propose a concept called e-predictability that quantifies the size of a model class ... Full text Cite

Detecting structural changes in dependent data

Conference 2017 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2017 - Proceedings · March 7, 2018 In the era of big data, a frequently encountered task is to model and identify structural changes in the data generating process. It is quite challenging especially when data are dependent and massive, requiring computationally efficient analysis. To addre ... Full text Cite

Modeling nonlinearity in multi-dimensional dependent data

Conference 2017 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2017 - Proceedings · March 7, 2018 Given massive data that may be time dependent and multi-dimensional, how to efficiently explore the underlying functional relationships across different dimensions and time lags? In this work, we propose a methodology to sequentially and adaptively model n ... Full text Cite

CLASSIFICATION OF LOCAL FIELD POTENTIALS USING GAUSSIAN SEQUENCE MODEL

Conference 2018 IEEE STATISTICAL SIGNAL PROCESSING WORKSHOP (SSP) · 2018 Link to item Cite

The number of independent sets in hexagonal graphs

Conference IEEE International Symposium on Information Theory Proceedings · August 9, 2017 We derive the tightest known bounds on η = 2ν, where ν is the growth rate of the logarithm of the number of independent sets on a hexagonal lattice. To obtain these bounds, we generalize a method proposed by Calkin and Wilf. Their original strat ... Full text Cite

Pseudo-wigner matrices from dual BCH codes

Conference IEEE International Symposium on Information Theory Proceedings · August 9, 2017 We consider the problem of generating symmetric pseudo-random sign (±1) matrices based on the similarity of their spectra to Wigner's semicircular law. We introduce r-independent pseudo-Wigner ensembles and prove closeness of their spectra to the semicircu ... Full text Cite

Explicit symmetric pseudo-random matrices

Conference IEEE International Symposium on Information Theory Proceedings · July 2, 2017 We consider the problem of generating symmetric pseudo-random sign (±1) matrices based on the similarity of their spectra to Wigner's semicircular law. Using binary m-sequences (Golomb sequences) of lengths n = 2m - 1, we give a simple explicit ... Full text Cite

Analysis of a privacy-preserving PCA algorithm using random matrix theory

Conference 2016 IEEE Global Conference on Signal and Information Processing Globalsip 2016 Proceedings · April 19, 2017 To generate useful summarization of data while maintaining privacy of sensitive information is a challenging task, especially in the big data era. The privacy-preserving principal component algorithm proposed in [1] is a promising approach when a low rank ... Full text Cite

Robustness analysis for an online decentralized descent power allocation algorithm

Conference 2016 Information Theory and Applications Workshop Ita 2016 · March 27, 2017 As independent service providers shift from conventional energy to renewable energy sources, the power distribution system will likely experience increasingly significant fluctuation in supply, given the uncertain and intermittent nature of renewable sourc ... Full text Cite

Dictionary learning and sparse coding-based denoising for high-resolution task functional connectivity MRI analysis

Conference Lecture Notes in Computer Science Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics · January 1, 2017 We propose a novel denoising framework for task functional Magnetic Resonance Imaging (tfMRI) data to delineate the high-resolution spatial pattern of the brain functional connectivity via dictionary learning and sparse coding (DLSC). In order to address t ... Full text Cite

Scaling laws and phase transitions for target detection in MIMO radar

Conference 2016 IEEE Information Theory Workshop Itw 2016 · October 21, 2016 The performance of MIMO radar has been a subject of intense study in the past decades. For such a system, however, the important phenomenon of phase transition has received little attention in the literature. In this paper, we study the phase transition on ... Full text Cite

Convergence of limited communications gradient methods

Conference Proceedings of the American Control Conference · July 28, 2016 Distributed control and decision making increasingly play a central role in economical and sustainable operation of cyber-physical systems. Nevertheless, the full potential of the technology has not yet been fully exploited in practice due to communication ... Full text Cite

Transmit beamforming in Rayleigh product MIMO channels: Ergodic mutual information and symbol error rate

Conference 2016 IEEE International Conference on Communications Icc 2016 · July 12, 2016 In this paper, we consider MIMO beamforming in the presence of Rayleigh product channels. Based on a derived largest eigenvalue distribution, the key performance metrics of the beamforming system are obtained, assuming perfect channel knowledge at the tran ... Full text Cite

Base camp quality of life standardization and improvement

Conference Proceedings International Carnahan Conference on Security Technology · July 2, 2016 The United States (US) Army has over 66,000 soldiers engaged in contingency operations across the world. Current budgetary constraints and an uncertain global security environment require these operations to be executed as efficiently as possible. Base cam ... Full text Cite

An adaptive localization technique for wireless capsule endoscopy

Conference International Symposium on Medical Information and Communication Technology Ismict · June 23, 2016 Wireless capsule endoscopy (WCE) is an emerging technique to enhance Gastroenterologists information about the patient's gastrointestinal (G.I.) tract. Localization of capsule inside human body in this case is an active area of research. This can be though ... Full text Cite

Distributed resource allocation using one-way communication with applications to power networks

Conference 2016 50th Annual Conference on Information Systems and Sciences Ciss 2016 · April 26, 2016 Typical coordination schemes for future power grids require two-way communications. Since the number of end power-consuming devices is large, the bandwidth requirements for such two-way communication schemes may be prohibitive. Motivated by this observatio ... Full text Cite

Data-driven learning of the number of states in multi-state autoregressive models

Conference 2015 53rd Annual Allerton Conference on Communication Control and Computing Allerton 2015 · April 4, 2016 In this work, we consider the class of multi-state autoregressive processes that can be used to model non-stationary time series of interest. In order to capture different autoregressive (AR) states underlying an observed time series, it is crucial to sele ... Full text Cite

Order Selection of Autoregressive Processes Using Bridge Criterion

Conference Proceedings 15th IEEE International Conference on Data Mining Workshop Icdmw 2015 · January 29, 2016 A new criterion is introduced for determining the order of an autoregressive model fit to time series data. The proposed technique is shown to give a consistent and asymptotically efficient order estimation. It has the benefits of the two well-known model ... Full text Cite

Learning the Number of Autoregressive Mixtures in Time Series Using the Gap Statistics

Conference Proceedings 15th IEEE International Conference on Data Mining Workshop Icdmw 2015 · January 29, 2016 Using a proper model to characterize a time series is crucial in making accurate predictions. In this work we use time-varying autoregressive process (TVAR) to describe non-stationary time series and model it as a mixture of multiple stable autoregressive ... Full text Cite

Sequential learning of Multi-state autoregressive time series

Conference Proceeding of the 2015 Research in Adaptive and Convergent Systems Racs 2015 · October 9, 2015 Modeling and forecasting streaming data has fundamental importance in many real world applications. In this paper, we present an online model selection technique that can be used to model non-stationary time series in a sequential manner. Multi-state autor ... Full text Cite

Peak sidelobe level gumbel distribution for arrays of randomly placed antennas

Conference IEEE National Radar Conference Proceedings · June 22, 2015 Extreme Value Theory (EVT) is used to analyze the peak sidelobe level distribution for array element positions with arbitrary probability distributions. Computations are discussed in the context of linear antenna arrays using electromagnetic energy. The re ... Full text Cite

Compressed sensing techniques for image reconstruction in optical interferometry

Conference Optics InfoBase Conference Papers · January 1, 2014 We develop a novel method of image reconstruction from bispectrum and magnitude observables of an optical interferometer using Compressed-Sensing techniques.We validate our method in simulation and with actual measurements from a Fizeau interferometer. © 2 ... Cite

The design of environmentally friendly networks using coordinated multi-point (CoMP) transmission

Conference 2014 IEEE 11th Consumer Communications and Networking Conference Ccnc 2014 · January 1, 2014 Full text Cite

An environmentally friendly approach to the relay-aided cellular network architecture

Conference 2014 IEEE 11th Consumer Communications and Networking Conference Ccnc 2014 · January 1, 2014 We solve the problem of environmentally friendly (i.e., energy-efficient) base station and relay-station planning for a simplified cooperative cellular network. We assume a two-hop transmission model based on the time-division multiplexing operation. We pr ... Full text Cite

Compressed sensing techniques for image reconstruction in optical interferometry

Conference Optics Infobase Conference Papers · January 1, 2014 We develop a novel method of image reconstruction from bispectrum and magnitude observables of an optical interferometer using Compressed-Sensing techniques.We validate our method in simulation and with actual measurements from a Fizeau interferometer. © 2 ... Full text Cite

Distortion-based achievability conditions for joint estimation of sparse signals and measurement parameters from undersampled acquisitions

Conference IEEE International Symposium on Information Theory Proceedings · January 1, 2013 In this paper, we consider an undersampling system model of the form y = A(T(x, θ)) + n, where x is a k-sparse signal, T(·, · is a (possibly non-linear) function specified by a parameter vector θ and acting on x, A is a sensing matrix, and n is additive me ... Full text Cite

Energy-efficient base-station topologies for green cellular networks

Conference 2013 IEEE 10th Consumer Communications and Networking Conference Ccnc 2013 · January 1, 2013 We consider the problem of energy-efficient base-station (BS) planning for green cellular network design. There exist a number of criteria for greenness in the literature, but we focus only on the energy-normalized throughput. We first model energy consump ... Full text Cite

Capacity analysis for Gaussian channels with memoryless nonlinear hardware

Conference IEEE International Conference on Communications · January 1, 2013 This paper presents the analysis of the impact of circuit nonlinearities on the transmission rate of a communication system. We consider a general transmission model that includes transmitter- and receiver-side nonlinearities, environmental thermal noise a ... Full text Cite

Semidefinite programming-based localization algorithm in networks with inhomogeneous media

Conference Proceeding of the 2012 ACM Research in Applied Computation Symposium Racs 2012 · December 1, 2012 In this paper, we study the asymptotic properties of a semidefinite programming (SDP) based localization algorithm in a network with inhomogeneous RF transmission medium given incomplete and inaccurate pairwise distance measurements between sensorssensors ... Full text Cite

Vertical spectrum sharing via distributed asynchronous interference avoidance in cognitive radio networks

Conference 2012 IEEE Consumer Communications and Networking Conference Ccnc 2012 · January 1, 2012 In this paper, we study the problem of distributed spectrum allocation under a vertical spectrum sharing scenario in a cognitive radio network. The secondary users share the spectrum licensed to the primary user by observing the activity statistics of the ... Full text Cite

Energy harvesting cooperative communications

Conference IEEE International Symposium on Personal Indoor and Mobile Radio Communications PIMRC · January 1, 2012 In this paper, we propose a new cooperative wireless transmission in a scenario where the source salvages the energy during the relay's transmission considering the fact that the source does not need to retrieve the transmitted message. We also evaluate a ... Full text Cite

Spectral distribution of the product of two random matrices based on binary block codes

Conference 2011 49th Annual Allerton Conference on Communication Control and Computing Allerton 2011 · December 1, 2011 In this paper, we study the spectral distribution of the product of two random matrices based on binary block codes, and prove that if the dual distances of the underlying codes are large enough, the asymptotic spectral distribution will be close to a dete ... Full text Cite

Jointly optimized two-cell MIMO systems

Conference 2011 IEEE Globecom Workshops GC Wkshps 2011 · December 1, 2011 Jointly optimized interference aware coordinated beamforming (IA-CBF) algorithms for a two-cell system are investigated, where each base/mobile station is equipped with multiple transmit antennas. We first derive a minimum-mean-square-error IA-CBF algorith ... Full text Cite

Regularized recursive least squares for anomaly detection in sparse channel tracking applications

Conference Proceedings of the 2011 ACM Research in Applied Computation Symposium Racs 2011 · December 1, 2011 In this paper, we study the problem of anomaly detection in sparse channel tracking applications via the l 1-regularized least squares adaptive filter (SPARLS). Anomalies arise due to unexpected adversarial changes in the channel and quick detec ... Full text Cite

Robust scalable physical layer network coding

Conference 2011 49th Annual Allerton Conference on Communication Control and Computing Allerton 2011 · December 1, 2011 We present a new method of physical layer network coding that transforms the received signal at a relay node before re-transmission by eliminating the unreliable components of the signal. This approach improves the physical layer network coding performance ... Full text Cite

Localization of mobile nodes based on inaccurate round-trip-time measurements using Bayesian inference

Conference Proceedings of the 2011 ACM Research in Applied Computation Symposium Racs 2011 · December 1, 2011 In this paper, we present a Bayesian-based localization method to estimate the unknown location of a mobile station (MS) in a network given the known location of a set of base stations. In this study, the MS was covered by a base station if the signal to i ... Full text Cite

Group randomness properties of pseudo-noise and gold sequences

Conference 12th Canadian Workshop on Information Theory Cwit 2011 · July 7, 2011 In this paper, we study the group randomness of pseudo-random sequences based on shortened first-order Reed-Muller codes and the Gold sequences. In particular, we characterize the empirical spectral distribution of random matrices from shortened first-orde ... Full text Cite

Optimum cognitive radio transmission scheme for reducing average interference power

Conference 12th Canadian Workshop on Information Theory Cwit 2011 · July 7, 2011 Reducing interference induced by secondary users is one of the most challenging problems in spectrum sharing. In this paper, we derive the optimum transmission strategy that minimizes the average interference power created by cognitive radios at a given po ... Full text Cite

Network error correction from matrix network coding

Conference 2011 Information Theory and Applications Workshop Ita 2011 Conference Proceedings · May 13, 2011 We present matrix network coding methods that are naturally amenable to a distributed implementation method, i.e., do not require the knowledge of network topology, and that are suitable for network error correction. First, the Singleton bound can be K-fol ... Full text Cite

Sidelobe level distribution computation for antenna arrays with arbitrary element distributions

Conference Conference Record Asilomar Conference on Signals Systems and Computers · January 1, 2011 This paper discusses the conditions in which the sidelobe level distribution can be calculated in closed form for arrays with randomly located antenna elements with arbitrary element probability distributions. By using the number of times a beampattern cro ... Full text Cite

Service coverage for cognitive radio networks with cooperative relays in shadowed hotspot areas

Conference 2011 IEEE Wireless Communications and Networking Conference Wcnc 2011 · January 1, 2011 In this paper, we investigate the service coverage for interweaving cognitive radio networks with cooperative relays in shadowed areas within the metropolitan region. We highlight the influence of relays on the primary and secondary systems in severe shado ... Full text Cite

Information-theoretic limits of dense underwater networks

Conference Conference Record Asilomar Conference on Signals Systems and Computers · January 1, 2011 Information-theoretic throughput scaling laws are analyzed in an underwater acoustic network with n regularly located nodes on a unit square, in which both bandwidth and received signal power can be severely limited. A narrow-band model is assumed where th ... Full text Cite

Low peak to average power ratio Turbo Block QPSK coded OFDM

Conference 2011 IEEE Consumer Communications and Networking Conference Ccnc 2011 · January 1, 2011 In this work, we propose a QPSK OFDM code that has both low peak to average power ratio and good bit error rate performance. We consider a Time-frequency Turbo Block code. This code is composed of the cosets of the first-order binary Reed-Muller (RM) code ... Full text Cite

Blind identification of sparse channels and symbol detection via the em algorithm

Conference 17th DSP 2011 International Conference on Digital Signal Processing Proceedings · January 1, 2011 In this paper, we address the problem of blind identification of sparse channels. For this purpose, the Expectation-Maximization is modified to accommodate channel sparsity. The resulting algorithm is applicable for linear and nonlinear channels. Computer ... Full text Cite

On the performance estimate for Turbo codes

Conference 2010 International Conference on Wireless Communications and Signal Processing Wcsp 2010 · December 1, 2010 Recently, the scaling law has been extended to approximate the waterfall performance for Turbo codes over the binary erasure channel. However, the waterfall approximation over other channels, such as the AWGN and Rayleigh fading channels, is still unknown ... Full text Cite

Analysis of network coded HARQ for multiple unicast flows

Conference IEEE International Conference on Communications · August 13, 2010 In this paper, we consider network coded (NCed) Hybrid-ARQ (HARQ) for multiple unicast flows. The main contribution of the paper is the derivation of throughput expression for NCed HARQ with arbitrary number of users in i.i.d. channels. We apply the result ... Full text Cite

Random frames from binary linear block codes

Conference 2010 44th Annual Conference on Information Sciences and Systems Ciss 2010 · June 24, 2010 Let C be an [n, k, d] binary linear block code of length n, dimension k and minimum Hamming distance d over GF(2n). Let d⊥ denote the minimum Hamming distance of the dual code of Cover GF(2n). Let ε : GF(2n) → {- ... Full text Cite

An exponential model for evaluating error performance of turbo codes

Conference Globecom IEEE Global Telecommunications Conference · January 1, 2010 In this paper, an exponential model is proposed to approximate the error performance of Turbo codes within low and intermediate SNR region. Our analysis unveils that, the proposed exponential model offers a simple but effective method to approximate the wa ... Full text Cite

Multi-hop routing is order-optimal in underwater extended networks

Conference IEEE International Symposium on Information Theory Proceedings · January 1, 2010 Capacity scaling laws are analyzed in an underwater acoustic network with n regularly located nodes. A narrowband model is assumed where the carrier frequency is allowed to scale as a function of n. In the network, we characterize an attenuation parameter ... Full text Cite

Low density frames for compressive sensing

Conference ICASSP IEEE International Conference on Acoustics Speech and Signal Processing Proceedings · January 1, 2010 We consider the compressive sensing of a sparse or compressible signal x ∈ ℝM. We explicitly construct a class of measurement matrices, referred to as the low density frames, and develop decoding algorithms that produce an accurate estimate x̂ ev ... Full text Cite

Adaptive algorithms for sparse nonlinear channel estimation

Conference IEEE Workshop on Statistical Signal Processing Proceedings · December 25, 2009 In this paper, we consider the estimation of sparse nonlinear communication channels. Transmission over the channels is represented by sparse Volterra models that incorporate the effect of Power Amplifiers. Channel estimation is performed by compressive se ... Full text Cite

TAS protocols of a PASD system with limited feedback information

Conference Globecom IEEE Global Telecommunications Conference · December 1, 2009 In a Protocol Assisted Switched Diversity (PASD) system, the time separated repetition blocks are accumulated in order to increase decoding reliability. In this paper, we consider transmit antenna selection (TAS) protocols in a PASD system. To minimize sys ... Full text Cite

Adaptive modulation and network coding with optimized precoding in two-way relaying

Conference Globecom IEEE Global Telecommunications Conference · December 1, 2009 We propose a precoding strategy which controls amplitude and phase of receiving signals to improve throughput for two-stage bidirectional relaying. We consider the case when the nodes know channel state information (CSI) and can adopt adaptive modulation t ... Full text Cite

An adaptive rate assignment strategy for CDMA2000 IS-856 subject to RAB delay

Conference Globecom IEEE Global Telecommunications Conference · December 1, 2009 In this paper, the problem of resource allocation in IS-856 uplink in the presence of time-delay is studied. A set of nonlinear adaptive controllers are designed to stabilize the wireless network and use the system resources efficiently. The controllers ob ... Full text Cite

Non-linear network coding in two-way relaying discrete channels

Conference 2009 International Conference on Wireless Communications and Signal Processing Wcsp 2009 · December 1, 2009 In this paper, we present optimal network coding which can maximize achievable rate region in two-way relaying discrete channels, where there are two communicating nodes and one relaying node. We focus on a low-complexity relaying protocol termed the detec ... Full text Cite

Cognitive networks achieve throughput scaling of a homogeneous network

Conference Final Proceedings of the 2009 7th International Symposium on Modeling and Optimization in Mobile Ad Hoc and Wireless Networks Wiopt 2009 · December 1, 2009 We study two distinct, but overlapping, networks which operate at the same time, space and frequency. The first network consists of n randomly distributed primary users, which form either an ad hoc network, or an infrastructure- supported ad hoc network in ... Full text Cite

Sphere packing optimization and EXIT chart analysis for multi-dimensional QAM signaling

Conference IEEE International Conference on Communications · November 19, 2009 We investigate on multi-dimensional QAM constellations optimized by sphere packing with the known densest lattices. We propose a greedy design method assisted by the sphere detection. It is demonstrated that the optimized constellations can significantly i ... Full text Cite

Denoising strategy for convolutionally-coded bidirectional relaying

Conference IEEE International Conference on Communications · November 19, 2009 In this paper, we present a forwarding strategy for two-stage bidirectional relaying in which trellis-coded modulation (TCM) is employed. We reveal that adaptive network coding cannot resolve distance shortening occurred at specific channel conditions when ... Full text Cite

Controller design for rate assignment in wireless networks

Conference IEEE International Conference on Communications · November 19, 2009 In this paper, data-rate assignment in IS-856 uplink (reverse link) is studied. The problem is first formulated in an interference model framework, and then a dynamic control strategy is developed for efficient rate assignment. In the first step, the contr ... Full text Cite

Interference analysis between on-the-move users and GEO satellites

Conference IEEE Vehicular Technology Conference · October 12, 2009 We consider on the move satellite systems with geostationary satellites. Using the antenna pointing distribution of Weerackody and Gonzalez, we derive tight upper and lower bounds of the interference between adjacent satellites from/to the on-the-move plat ... Full text Cite

Comparison of SPARLS and RLS algorithms for adaptive filtering

Conference 2009 IEEE Sarnoff Symposium Sarnoff 2009 Conference Proceedings · July 23, 2009 In this paper, we overview the Low Complexity Recursive ℒ1-Regularized Least Squares (SPARLS) algorithm proposed in [2], for the estimation of sparse signals in an adaptive filtering setting. The SPARLS algorithm is based on an Expectation-Maxim ... Full text Cite

A class of Bi-directional multi-relay protocols

Conference IEEE International Symposium on Information Theory Proceedings · January 1, 2009 In a bi-directional relay channel, two nodes wish to exchange independent messages over a shared wireless halfduplex channel with the help of relays. Recent work has considered information theoretic limits of the bi-directional relay channel with a single ... Full text Cite

Denoising Strategy for Convolutionally-Coded Bidirectional Relaying

Conference 2009 IEEE INTERNATIONAL CONFERENCE ON COMMUNICATIONS, VOLS 1-8 · January 1, 2009 Link to item Cite

TAS Protocols of a PASD system with limited feedback information

Conference GLOBECOM 2009 - 2009 IEEE GLOBAL TELECOMMUNICATIONS CONFERENCE, VOLS 1-8 · January 1, 2009 Link to item Cite

Controller Design for Rate Assignment in Wireless Networks

Conference 2009 IEEE INTERNATIONAL CONFERENCE ON COMMUNICATIONS, VOLS 1-8 · January 1, 2009 Link to item Cite

Adaptive Modulation and Network Coding with Optimized Precoding in Two-Way Relaying

Conference GLOBECOM 2009 - 2009 IEEE GLOBAL TELECOMMUNICATIONS CONFERENCE, VOLS 1-8 · January 1, 2009 Link to item Cite

Sphere Packing Optimization and EXIT Chart Analysis for Multi-Dimensional QAM Signaling

Conference 2009 IEEE INTERNATIONAL CONFERENCE ON COMMUNICATIONS, VOLS 1-8 · January 1, 2009 Link to item Cite

An Adaptive Rate Assignment Strategy for CDMA2000 IS-856 Subject to RAB Delay

Conference GLOBECOM 2009 - 2009 IEEE GLOBAL TELECOMMUNICATIONS CONFERENCE, VOLS 1-8 · January 1, 2009 Link to item Cite

The impact of spectrum sensing time on the performance of the GADIA algorithm

Conference 2008 IEEE Symposium on New Frontiers in Dynamic Spectrum Access Networks Dyspan 2008 · December 29, 2008 In this paper, we analyze the performance of the GADIA (Greedy Asynchronous Distributed Interference Avoidance) algorithm ([1], [2]) given the spectral sensing time of the various active agents. Under the assumption that the sensing time is small compared ... Full text Cite

Beam selection gain from butler matrices

Conference IEEE Vehicular Technology Conference · December 1, 2008 We consider a wireless transmission scenario, when a base station is endowed with a fixed beamforming network, where M antennas are employed at the base station to point beams to predetermined azimuthal angles. In our setting, a Butler matrix is deployed a ... Full text Cite

Interference reduction by beamforming in cognitive networks

Conference Globecom IEEE Global Telecommunications Conference · December 1, 2008 We consider beamforming in a cognitive network with multiple primary users and secondary users sharing the same spectrum. In particular, we assume that each secondary transmitter has Nt antennas and transmits data to its single- antenna receiver ... Full text Cite

Denoising maps and constellations for wireless network coding in two-way relaying systems

Conference Globecom IEEE Global Telecommunications Conference · December 1, 2008 We investigate on the design of modulation schemes suited for two-way wireless relaying systems that apply network coding at the physical layer. We consider network coding based on denoise-and-forward (DNF), which consists of two stages: multiple access (M ... Full text Cite

Ground-to-air interference analysis in cellular ATG systems

Conference Ciss 2008 the 42nd Annual Conference on Information Sciences and Systems · September 22, 2008 For air-to-ground cellular systems with no frequency reuse, we provide an analysis of the inter-cell interference under 3D hexagonal cell planning and a line of sight channel model with no shadowing and no fading assumptions. Based on this model, we provid ... Full text Cite

Noisy compressive sampling limits in linear and sublinear regimes

Conference Ciss 2008 the 42nd Annual Conference on Information Sciences and Systems · September 22, 2008 The authors have recently established a set of results that characterize the number of measurements required to recover a sparse signal in ℂM with L non-zero coefficients from compressed samples in the presence of noise. These results indicate t ... Full text Cite

A distributed asynchronous algorithm for spectrum sharing in wireless ad hoc networks

Conference Ciss 2008 the 42nd Annual Conference on Information Sciences and Systems · September 22, 2008 We propose a distributed asynchronous spectrum allocation algorithm that achieves performance close to that of a centralized optimal algorithm. In our network model, nodes are grouped into a number of clusters. Each cluster chooses its transmission frequen ... Full text Cite

On the degrees-of-freedom of the MIMO interference channel

Conference Ciss 2008 the 42nd Annual Conference on Information Sciences and Systems · September 22, 2008 The high signal-to-noise ratio capacity of the symmetric MIMO interference channel is characterized as a function of the interference-to-noise ratio. This work is a multiple antenna extension of the degrees of freedom expressions derived by Etkin et al. fo ... Full text Cite

Relay selection methods for wireless cooperative communications

Conference Ciss 2008 the 42nd Annual Conference on Information Sciences and Systems · September 22, 2008 We study the problem of relay selection in a wireless cooperative network. Assuming a single source, a single destination, and N uniformly distributed candidate relays, we seek to select a set of cooperating relays to minimize the total transmission time o ... Full text Cite

Interference in air-to-ground cellular systems

Conference IEEE International Conference on Communications · September 12, 2008 Interference in cellular terrestrial systems has been extensively studied in the literature mainly through the use of simulations. The interference for CDMA cellular air-to-ground systems has been similarly evaluated through extensive simulations. However, ... Full text Cite

Cognitive sensing based on side information

Conference Proceedings of the 2008 IEEE Sarnoff Symposium Sarnoff · August 14, 2008 We study a sensing algorithm for cognitive radios based on Bayesian energy detection while utilizing available side information. The side information available to the cognitive user can consist of: (i) spatial locations of the cognitive and primary receive ... Full text Cite

Comparison of bi-directional relaying protocols

Conference Proceedings of the 2008 IEEE Sarnoff Symposium Sarnoff · August 14, 2008 In a bi-directional relay channel, two nodes wish to exchange independent messages over a shared wireless channel with the help of a relay. In this paper, we derive achievable rate regions for four new half-duplex protocols and compare these to four existi ... Full text Cite

Distributed dynamic frequency allocation in wireless networks under time-varying user activities

Conference Proceedings of the 2008 IEEE Sarnoff Symposium Sarnoff · August 14, 2008 We propose a framework to analyze the performance of the distributed dynamic frequency allocation algorithm introduced in [1] in the presence of time-varying activity of clusters. In our network model, nodes are grouped into a number of clusters. Each clus ... Full text Cite

Iterative approach to base station positioning in cellular networks

Conference Proceedings of the 2008 IEEE Sarnoff Symposium Sarnoff · August 14, 2008 We consider the problem of finding the optimal spatial configuration of a number of base stations in a cellular network. We propose a framework to approximately maximize the average Signal to Interference plus Noise Ratio (SINR), with respect to the positi ... Full text Cite

On the peak to average power reduction of OFDM signals using reserved subcarriers

Conference IEEE International Symposium on Personal Indoor and Mobile Radio Communications PIMRC · January 1, 2008 Orthogonal Frequency Division Multiplexing (OFDM) is an attractive technique for high data rate transmissions in wireless and wireline systems. However OFDM signals suffer from high peak-to-mean envelope power (PMEPR). In this paper we consider the problem ... Full text Cite

Performance analysis of a PASD antenna system in Rayleigh fading channels

Conference IEEE Wireless Communications and Networking Conference Wcnc · January 1, 2008 In this paper, we analyze the performance of a communication system that employs Protocol Assisted Switched Diversity (PASD) antennas under Rayleigh fading channel conditions. The PASD system accumulates time-displaced blocks, each with the same informatio ... Full text Cite

Interference in a cognitive network with beacon

Conference IEEE Wireless Communications and Networking Conference Wcnc · January 1, 2008 We study a cognitive network consisting of multiple cognitive users communicating in the presence of a single primary user. The primary user is located at the center of the network, and the cognitive users are uniformly distributed within a circle around t ... Full text Cite

Improved throughput scaling in wireless ad hoc networks with infrastructure

Conference IEEE International Symposium on Information Theory Proceedings · January 1, 2008 We analyze the benefits of infrastructure support in improving the throughput scaling in networks of n randomly located wireless nodes. The infrastructure uses multi-antenna base stations (BSs), in which the number of BSs and the number of antennas at each ... Full text Cite

The primary exclusive region in cognitive networks

Conference 2008 5th IEEE Consumer Communications and Networking Conference Ccnc 2008 · January 1, 2008 In this paper, we consider a cognitive network in which a single primary transmitter communicates with primary receivers within an area of radius R 0, called the primary exclusive region (PER). Inside this region, no cognitive users may transmit ... Full text Cite

Two-way relaying with network coding for frequency-selective fading channels

Conference Conference Record Asilomar Conference on Signals Systems and Computers · January 1, 2008 In this paper, we present a design method of physical layer network coding for reliable two-way relaying in frequency- selective fading channels. We consider a two-stage protocol: At the first stage, two terminals simultaneously transmit data packets towar ... Full text Cite

On the Peak to Average Power Reduction of OFDM Signals Using Reserved Subcarriers

Conference 2008 IEEE 19TH INTERNATIONAL SYMPOSIUM ON PERSONAL, INDOOR AND MOBILE RADIO COMMUNICATIONS · January 1, 2008 Link to item Cite

Beam Selection Gain from Butler Matrices

Conference 68TH IEEE VEHICULAR TECHNOLOGY CONFERENCE, FALL 2008 · January 1, 2008 Link to item Cite

Performance study of various sparse representation methods using redundant frames

Conference Forty First Annual Conference on Information Sciences and Systems Ciss 2007 Proceedings · December 1, 2007 Sparse representations have recently received wide attention because of their numerous potential applications. In this paper, we consider sparse representations of signals with at most L non-zero coefficients using a frame ℱ of size M in CN. We ... Full text Cite

Antenna selection strategies for MIMO systems

Conference IEEE International Conference on Communications · December 1, 2007 Antenna selection strategies for multiple-input multiple-output (MIMO) wireless systems is presented. In our scheme, both transmit and receive antenna selection is done at the receiver, significantly reducing feedback information to the transmitter. Row/co ... Full text Cite

A new practical dirty-paper coding strategy in MIMO system

Conference Forty First Annual Conference on Information Sciences and Systems Ciss 2007 Proceedings · December 1, 2007 In this paper, we propose a new broadcast strategy for a MIMO system with N transmit antennas at the transmitter and M < N single antenna receivers. The proposed method spatially separates the M users but does not suffer from the power loss of classical SD ... Full text Cite

Coded bi-directional relaying in combat scenarios

Conference Proceedings IEEE Military Communications Conference MILCOM · December 1, 2007 In this paper, we present three protocols for bidirectional coded cooperation. In coded bi-directional cooperation, two agents wish to communicate with each other in the presence of a relay node that may help. The first protocol is a two phase protocol whe ... Full text Cite

Adaptive frequency synchronization for collaborative communication systems

Conference Proceedings International Conference on Distributed Computing Systems · January 1, 2007 Cooperative diversity systems are wireless communication systems designed to exploit cooperation among users to mitigate the effects of multipath fading. In fairly general conditions, it has been shown that these systems can achieve the diversity order of ... Full text Cite

Scaling Laws of Cognitive Networks.

Conference CrownCom · 2007 Cite

Training sequence design for wireless collaborative communication systems in frequency-selective fading

Conference Conference Record Asilomar Conference on Signals Systems and Computers · January 1, 2007 We consider a collaborative communication system consisting of several nodes jointly communicating to a common destination node. Once each node determines the message to send, a space-time code across the sources is used to transmit the signal to the desti ... Full text Cite

On sparsity, redundancy and quality of frame representations

Conference IEEE International Symposium on Information Theory Proceedings · January 1, 2007 We consider approximations of signals by the elements of a frame in a complex vector space of dimension N and formulate both the noiseless and the noisy sparse representation problems. The noiseless representation problem is to find sparse representations ... Full text Cite

Channel hardening and the scheduling gain of antenna selection diversity schemes

Conference IEEE International Symposium on Information Theory Proceedings · January 1, 2007 For a multiple antenna system, we compute the asymptotic distribution of antenna selection gain when the transmitter selects the transmit antenna with the strongest channel. We use this to asymptotically estimate the underlying channel capacity distributio ... Full text Cite

Distributed delay-aware rate control algorithm for the reverse-link of CDMA systems

Conference IEEE International Conference on Communications · January 1, 2006 We propose a new distributed rate control algorithm for the reverse link of CDMA systems. Specifically, when the statistics of packet arrivals are known locally to each access terminal (AT), this new algorithm aims to provide a means to control the relativ ... Full text Cite

Differential spatial multiplexing for two and three transmit antennas

Conference IEEE International Conference on Communications · January 1, 2006 In this paper, we construct a differential spatial multiplexing method that can trade-off spatial diversity for increased transmission rate in differential multi-input multi-output (MIMO) systems. To achieve a desired multiplexing gain while providing simp ... Full text Cite

Cognitive decomposition of wireless networks

Conference 1st International Conference on Cognitive Radio Oriented Wireless Networks and Communications 2006 Crowncom · January 1, 2006 In this paper, we provide a framework for a fundamental study of the communication limits of networks of cognitive devices. It is shown that all communication in a network of cognitive and non-cognitive devices can be cast into competitive, cognitive and c ... Full text Cite

Cognitive decomposition of wireless networks

Conference 2006 1ST INTERNATIONAL CONFERENCE ON COGNITIVE RADIO ORIENTED WIRELESS NETWORKS AND COMMUNICATIONS · January 1, 2006 Link to item Cite

Performance evaluation of MB-OFDM and DS-UWB systems for wireless personal area networks

Conference Icu 2005 2005 IEEE International Conference on Ultra Wideband Conference Proceedings · December 1, 2005 Ultra-wideband (UWB) radio has been proposed for physical layer standard of the future high-speed wireless personal area networks (WFANs), One proposal is referred to as multi-band orthogonal frequency division multiplexing (MB-OFDM), and the other is dire ... Cite

Trading-off transmission rate with transmit diversity in differential detection

Conference IEEE Vehicular Technology Conference · October 17, 2005 In this paper we construct a method where rate can be trade-off for diversity in differential detection transmit diversity. Specifically, we show for the case 3 and 4 transmit and M receive anteanns, when using the PSK constellation, it is possible to doub ... Cite

Delay bounds for packetizing time-varying fluid policies with speedup and lookahead in single server systems

Conference Proceedings IEEE INFOCOM · October 7, 2005 We consider an online, processor sharing system with a single server in a packet-switched network where the processor multiplexes packets from multiple input ports onto the same output link. Under a fluid model, in which incoming data is treated as infinit ... Full text Cite

Optimum opportunistic beamforming based on multiple weighting vectors

Conference IEEE International Conference on Communications · September 13, 2005 In order to improve the throughput of the opportunistic beamforming, we generalize the opportunistic beamforming by using multiple random weighting vectors at each time slot. The base station chooses the best weighting vector and performs the opportunistic ... Cite

Collaborative diversity enhancements for wireless communications

Conference IEEE International Conference on Communications · September 12, 2005 The use of the spatial dimension is known to greatly increase the reliability of quasi-static (non-ergodic) wireless channels. In this paper, we demonstrate that most of this gain can also be achieved through collaborative communications with single-antenn ... Cite

On the effects of phase estimation errors on collaborative beamforming in wireless ad hoc networks

Conference ICASSP IEEE International Conference on Acoustics Speech and Signal Processing Proceedings · January 1, 2005 The performance of collaborative beamforming is studied using the theory of random arrays in the framework of wireless sensor networks. With the application to ad hoc networks in mind, two scenarios, one denoted closed-loop and the other open-loop, are con ... Full text Cite

Cognitive multiple access networks

Conference IEEE International Symposium on Information Theory Proceedings · January 1, 2005 A cognitive radio can sense the transmission of other users in its environment and possibly extract the corresponding messages. It can use this information to transmit over the same channel while reducing interference from, and to other users. In this pape ... Full text Cite

A signal processing application of randomized low-rank approximations

Conference 2005 IEEE/SP 13TH WORKSHOP ON STATISTICAL SIGNAL PROCESSING (SSP), VOLS 1 AND 2 · January 1, 2005 Link to item Cite

A signal processing application of randomized low-rank approximations

Conference IEEE Workshop on Statistical Signal Processing Proceedings · January 1, 2005 Low-rank approximations to linear operators find wide use in signal processing. In the discrete case, assuming the desired rank is known a priori, such approximations are generally calculated using the singular value decomposition. In this vein, randomized ... Full text Cite

Design and analysis of collaborative diversity protocols for wireless sensor networks

Conference IEEE Vehicular Technology Conference · December 1, 2004 We consider wireless sensor networks where all the sensor nodes share the same channel and transmit collaboratively in a quasi-static Rayleigh fading environment. Intuition suggests that collaborative communication can achieve a higher diversity gain than ... Cite

Time-hopping sequence design for narrowband interference suppression

Conference IEEE Vehicular Technology Conference · December 1, 2004 In this paper, we present a simple interference mitigation solution for coexistence of Ultra-Wideband (UWB) systems with other wireless systems. Specifically, we consider the design of time-hopping codes for UWB systems that deploy impulse radio architectu ... Cite

Collaborative beamforming in ad hoc networks

Conference 2004 IEEE Information Theory Workshop Proceedings Itw · December 1, 2004 The performance of collaborative beamforming is analyzed using the theory of random arrays. The statistical average and distribution of the beampatterns of randomly generated phased arrays are derived in the framework of wireless ad hoc sensor networks. Ea ... Cite

Universal bounds on buffer size for packetizing fluid policies in input queued, crossbar switches

Conference Proceedings IEEE INFOCOM · November 22, 2004 In this paper, we consider a type of on-line, traffic scheduling problem in input queued, crossbar switches. The input to a problem, at each time step, is a set of desired traffic rates. These traffic rates in general cannot be exactly achieved since they ... Full text Cite

Frequency synchronization in OFDM

Conference ICASSP IEEE International Conference on Acoustics Speech and Signal Processing Proceedings · September 27, 2004 We present an OFDM frequency synchronization scheme. The scheme uses periodic OFDM symbols, similarly to the algorithms proposed previously by Morelli and Mengali and Schmidl and Cox. The proposed scheme attains considerably higher accuracy than the scheme ... Cite

Differential space time block codes using QAM for four transmit antennas

Conference IEEE International Conference on Communications · August 30, 2004 We develop differential space time block codes (STBC) using non-constant modulus constellations for four transmit antennas. The proposed method improves on the conventional differential STBC techniques because of the larger minimum distances of non-constan ... Cite

An Impulse Response Model For Residential Wireless Channels

Conference Conference Record IEEE Global Telecommunications Conference · December 1, 2003 We present a statistical model for the Multipath Intensity Profile (MIP) that can be used to construct an impulse response of a residential wireless channel. The model incorporates relevant parameters such as distance, global rms delay spread and the corre ... Cite

Coexistence of Ultra-Wideband Systems with IEEE-802.1la Wireless LANs

Conference Conference Record IEEE Global Telecommunications Conference · December 1, 2003 In this study we provide a physical layer based analysis of the coexistence issues of Ultra-Wideband (UWB) with other devices in the same spectrum. Specifically, we have focused on the UWB interference to and from devices using the Wireless Local Area Netw ... Cite

Complementary beamforming

Conference IEEE Vehicular Technology Conference · December 1, 2003 We present complementary beamforming techniques which enable the formation of multiple directional beams to some users while guaranteeing a minimum power level to the rest of the users in the network. These techniques can be used for many different applica ... Cite

UWB path loss characterization in residential environments

Conference IEEE Radio Frequency Integrated Circuits Symposium RFIC Digest of Technical Papers · August 22, 2003 In this paper, we describe a simple method for measurement of the Ultra-Wideband Band (UWB) frequency response for evaluation of the path loss and impulse response of the UWB indoor channel. We propose a simple statistical path loss model for the residenti ... Cite

UWB path loss characterization in residential environments

Conference IEEE MTT S International Microwave Symposium Digest · August 18, 2003 In this paper, we describe a simple method for measurement of the Ultra-Wideband Band (UWB) frequency response for evaluation of the path loss and impulse response of the UWB indoor channel We propose a simple statistical path loss model for the residentia ... Cite

A multipath intensity profile model for residential environments

Conference IEEE Wireless Communications and Networking Conference Wcnc · January 1, 2003 We describe a statistical model for the multipath intensity profile (MIP) in residential environments. The work is based on over 300,000 frequency response measurements at 712 locations in 23 homes. We show that, for non-line-of-sight (NLS) path, the MIP i ... Full text Cite

Differential space time block codes using QAM constellations

Conference IEEE International Symposium on Personal Indoor and Mobile Radio Communications PIMRC · January 1, 2003 We propose differential space time block codes (STBC) using quadrature amplitude modulation (QAM), which can not be utilized in the conventional differential STBC. Since QAM constellations have a larger minimum distance compared with phase shift keying (PS ... Full text Cite

UWB indoor path loss model for residential and commercial buildings

Conference IEEE Vehicular Technology Conference · January 1, 2003 We present a statistical model for the path loss of ultra-wideband channels in indoor environments. In contrast to previous measurements, the data reported here are for a bandwidth of 6 GHz rather than 1.25 GHz; they encompass commercial buildings in addit ... Full text Cite

Invited - UWB path loss characterization in residential environments

Conference 2003 IEEE MTT-S INTERNATIONAL MICROWAVE SYMPOSIUM DIGEST, VOLS 1-3 · January 1, 2003 Link to item Cite

UWB indoor delay profile model for residential and commercial environments

Conference IEEE Vehicular Technology Conference · January 1, 2003 We present a statistical model for the delay profile of ultra-wideband channels in indoor environments. Two kinds of profiles are defined, namely the multipath intensity profile (MIP) and the power delay profile (PDF). The MIP is he delay profile at a poin ... Full text Cite

Existence and construction of block interleavers

Conference IEEE International Conference on Communications · January 1, 2002 The necessary and sufficient conditions on the existence of (N1, N2) block interleavers of block size L, were established. The interleaver was found to exist on the condition that L ≥ N1N2. Explicit algebraic exp ... Full text Cite

Maximum achievable energy reduction using coding with applications to deep sub-micron buses

Conference Proceedings IEEE International Symposium on Circuits and Systems · January 1, 2002 In this work we ask: How much can we reduce the power consumption in deep-sub-micron buses using coding techniques? We answer the question in two steps. First we give the minimum energy per information bit required for communicating through deep-sub-micron ... Cite

Autoregressive modeling of an indoor UWB channel

Conference 2002 IEEE Conference on Ultra Wideband Systems and Technologies Uwbst 2002 Digest of Papers · January 1, 2002 Based on frequency domain measurements in the 4.375-5.625 GHz band a channel model for the frequency response of the indoor radio channel is introduced. In particular. a second-order Autoregressive (AR) model is proposed for frequency response generation o ... Full text Cite

A statistical path loss model for in-home UWB channels

Conference 2002 IEEE Conference on Ultra Wideband Systems and Technologies Uwbst 2002 Digest of Papers · January 1, 2002 This paper describes a simple statistical model for evaluating the path loss in residential environments. It consists of detailed characterization of path loss model parameters of ultra-wideband band (UWB) signals having a nominal center frequency of 5 GHz ... Full text Cite

Two constructions of 16-QAM Golay complementary sequences

Conference IEEE International Symposium on Information Theory Proceedings · January 1, 2002 Two constructions of 16-quadrature amplitude modulation (QAM) Golay complementary sequences from 4-QAM complementary sequences were presented. These sequences can be applied as pilot sequences in future orthogonal frequency division miltiplexing (OFDM) sys ... Cite

Variable rate sapce-time block codes in MPSK systems

Conference Conference Record of the Asilomar Conference on Signals Systems and Computers · January 1, 2002 We consider a multiple antenna system when combined array processing with space-time coding is used [2]. We present variable rate space-time block codes for two, three, and four transmit antennas and optimize the transmit power so that the average BER is m ... Cite

Energy reduction and fundamental energy limits in digital VLSI circuits

Conference IEEE International Symposium on Information Theory Proceedings · January 1, 2002 Energy reduction and fundamental energy limits in digital VLSI circuits were discussed. Bus was used to calculate the minimum cost of energy since its cost function is known. Results can be extended to units modelled as finite state machines and can also b ... Cite

A channel-shortening multiuser detector for DS-CDMA systems

Conference IEEE Vehicular Technology Conference · January 1, 2001 The multiple-access interference (MAI) encountered by Direct-Sequence Code-Division Multiple Access (DS-CDMA) users in a multipath environment is a key issue in the detection of mobile communication users. The MAI is one of the major reasons for the degrad ... Cite

A simple encodable/decodable OFDM QPSK code with low peak-to-mean envelope power ratio

Conference IEEE International Symposium on Information Theory Proceedings · January 1, 2001 We present a QPSK coding scheme for use in multicarrier communications. The scheme supports simple encoding and extremely simple maximum likelihood (ML) decoding algorithms for orthogonal frequency division multiplexing (OFDM) applications. Moreover, all t ... Full text Cite

Image clustering using fuzzy graph theory

Conference Proceedings of SPIE the International Society for Optical Engineering · January 1, 2000 We propose an image clustering algorithm which uses fuzzy graph theory. First, we define a fuzzy graph and the concept of connectivity for a fuzzy graph. Then, based on our definition of connectivity we propose an algorithm which finds connected subgraphs ... Cite

On the existence and construction of good codes with low peak-to-average power ratios

Conference IEEE International Symposium on Information Theory Proceedings · January 1, 2000 The peak-to-average power ratio PAPR(C) of a code C is an important characteristic of that code when it is used in OFDM communications. We establish bounds on the region of achievable triples (R, d, PAPR(C)) where R is the code rate and d is the minimum Eu ... Full text Cite

A differential detection scheme for transmit diversity

Conference IEEE Wireless Communications and Networking Conference Wcnc · January 1, 1999 We present a transmission scheme for exploiting diversity given by two transmit antennas when neither the transmitter nor the receiver has access to channel state information. The new detection scheme can use PSK constellations and encoding is simple. At t ... Full text Cite

Space-time block coding-from a physical perspective

Conference IEEE Wireless Communications and Networking Conference Wcnc · January 1, 1999 Space-time block coding (STBC) is an effective transmit diversity scheme. It was presented previously by in terms of coding and decoding by Alamouti (see IEEE J. Select. Areas Commun., vol.16, no.8, p.1451-58, 1998) and by Tarokh, Jafarkhani and Calderbank ... Full text Cite

Algorithm for reducing the peak to average power ratio in a multicarrier communications system

Conference IEEE Vts 50th Vehicular Technology Conference Vtc 1999 Fall · January 1, 1999 One of the main implementation disadvantages of a multicarrier communication system is the possibly high peak to average power ratio of the transmitted signals. One proposed solution is given by Jones and Wilkinson. It is based on applying a specific phase ... Cite

Multiple description trellis coded quantization

Conference IEEE International Conference on Image Processing · December 1, 1998 We present a construction of multiple description trellis coded quantizers. We use the tensor product of trellises to build a trellis which is applicable to multiple description coding. The problems of index assignment and set partitioning for the resultin ... Cite

Space-time coded OFDM for high data-rate wireless communication over wideband channels

Conference IEEE Vehicular Technology Conference · December 1, 1998 The space-time coded orthogonal frequency division multiplexing (OFDM) scheme is described to provide high data-rate wireless communication over wideband channels. Simulation results for a 4-PSK 16 state space-time coded OFDM system are presented. The prop ... Cite

The application of orthogonal designs to wireless communication

Conference 1998 Information Theory Workshop Itw 1998 · January 1, 1998 We introduce space-block codzng, a new paradigm for transmission over Rayleigh fading channels using multiple transmit antennas. Data is encoded using a space-block code and the encoded data is split into n streams which are simultaneously transmitted usin ... Full text Cite

Successively refinable trellis coded quantization

Conference Data Compression Conference Proceedings · January 1, 1998 We propose successively refinable trellis coded quantizers which are suitable for progressive transmission. A new trellis structure which is scalable is used in the design of our trellis coded quantizers. A hierarchical set partitioning is used to preserve ... Cite

New detection schemes for transmit diversity with no channel estimation

Conference Icupc 1998 IEEE 1998 International Conference on Universal Personal Communications Conference Proceedings · January 1, 1998 We present two new detection methods for the simple transmit diversity scheme proposed by Alamouti (see IEEE Journal on Selected Areas of Communications, Special Issue on Signal Processing for Wireless Communications, 1998). The new detection schemes requi ... Full text Cite

Trellis-coded modulation and transmit diversity: Design criteria and performance evaluation

Conference Icupc 1998 IEEE 1998 International Conference on Universal Personal Communications Conference Proceedings · January 1, 1998 This paper presents the selection criteria for trellis codes used with the simple transmit diversity scheme proposed in Alamouti (1998) and reviewed in this paper. The optimum metrics for soft decision Viterbi decoding are derived and the asymptotic coding ... Full text Cite

Principal ratio combining for fixed wireless applications when transmitter diversity is employed

Conference Icupc 1998 IEEE 1998 International Conference on Universal Personal Communications Conference Proceedings · January 1, 1998 We present a near-optimal detection method for decoding codes appropriate to transmission using multiple transmit antennas in a fixed wireless environment. The method is a non-trivial generalization of maximum ratio combining. This reduces the decoding com ... Full text Cite

Space-time coded modulation for high data rate wireless communications

Conference Conference Record IEEE Global Telecommunications Conference · December 1, 1997 This paper presents the theory and practice of a new advanced modem technology suitable for high data rate wireless communications and presents its performance over a frequency-flat Rayleigh fading channel. The new technology is based on space-time coded m ... Cite

Upper bounds on trellis complexity of lattices

Conference IEEE International Symposium on Information Theory Proceedings · January 1, 1997 Unlike block codes, n-dimensional lattices can have minimal trellis diagrams with an arbitrarily large number of states, branches, and paths. We show by a counter-example that there is no f(n), a function of n, such that all rational lattices of dimension ... Full text Cite

Sequential decoding of lattice codes

Conference IEEE International Symposium on Information Theory Proceedings · January 1, 1997 We consider lattice tree-codes based on a lattice Λ having a finite trellis diagram T. Such codes are easy to encode and benefit from the structure of Λ . Sequential decoding of lattice tree-codes is studied, and the corresponding Fano (1963) metric is der ... Full text Cite

Trellis complexity versus the coding gain of lattices .2.

Conference IEEE TRANSACTIONS ON INFORMATION THEORY · November 1, 1996 Link to item Cite

Trellis complexity versus the coding gain of lattices

Conference IEEE International Symposium on Information Theory Proceedings · January 1, 1995 The growth of trellis diagrams of lattices versus their coding gain is studied. It is established that this growth exponentially in terms of the coding gain. ... Cite

Projected Latent Markov Chain Monte Carlo: Conditional Sampling of Normalizing Flows

Conference We introduce Projected Latent Markov Chain Monte Carlo (PL-MCMC), a technique for sampling from the exact conditional distributions learned by normalizing flows. As a conditional sampling method, PL-MCMC enables Monte Carlo Expectation Maximization (MC-EM) ... Cite

Task-Aware Neural Architecture Search”, International Conference on Acoustics, Speech, and Signal Processing

Conference The design of handcrafted neural networks requires a lot of time and resources. Recent techniques in Neural Architecture Search (NAS) have proven to be competitive or better than traditional handcrafted design, although they require domain knowledge and ha ... Cite