Carlo Tomasi

Conference Advances in Neural Information Processing Systems · January 1, 2025 Accurate and scalable quantification of animal pose and appearance is crucial for studying behavior. Current 3D pose estimation techniques, such as keypoint- and mesh-based techniques, often face challenges including limited representational detail, labor- ... Cite

Cross-Attention Transformer for Video Interpolation

Conference Lecture Notes in Computer Science · January 1, 2023 We propose TAIN (Transformers and Attention for video INterpolation), a residual neural network for video interpolation, which aims to interpolate an intermediate frame given two consecutive image frames around it. We first present a novel vision transform ... Full text Cite

SemARFlow: Injecting Semantics into Unsupervised Optical Flow Estimation for Autonomous Driving

Conference Proceedings of the IEEE International Conference on Computer Vision · January 1, 2023 Unsupervised optical flow estimation is especially hard near occlusions and motion boundaries and in low-texture regions. We show that additional information such as semantics and domain knowledge can help better constrain this problem. We introduce SemARF ... Full text Cite

Optical Flow Training Under Limited Label Budget via Active Learning

Conference Lecture Notes in Computer Science Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics · January 1, 2022 Supervised training of optical flow predictors generally yields better accuracy than unsupervised training. However, the improved performance comes at an often high annotation cost. Semi-supervised training trades off accuracy against annotation cost. We u ... Full text Cite

Unsupervised Flow Refinement near Motion Boundaries

Conference Bmvc 2022 33rd British Machine Vision Conference Proceedings · January 1, 2022 Unsupervised optical flow estimators based on deep learning have attracted increasing attention due to the cost and difficulty of annotating for ground truth. Although performance measured by average End-Point Error (EPE) has improved over the years, flow ... Cite

Joint Detection of Motion Boundaries and Occlusions

Conference 32nd British Machine Vision Conference Bmvc 2021 · January 1, 2021 We propose MONet, a convolutional neural network that jointly detects motion boundaries (MBs) and occlusion regions (Occs) in video both forward and backward in time. Detection is difficult because optical flow is discontinuous along MBs and undefined in O ... Cite

Person re-identification from gait using an autocorrelation network

Conference IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops · June 1, 2019 We propose a new biometric feature based on autocorrelation using an end-to-end trained network to capture human gait from different viewpoints. Our method condenses an unbounded image stream into a fixed size descriptor, and capitalizes on the periodic na ... Full text Cite

Features for Multi-target Multi-camera Tracking and Re-identification

Conference Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition · December 14, 2018 Multi-Target Multi-Camera Tracking (MTMCT) tracks many people through video taken from several cameras. Person Re-Identification (Re-ID) retrieves from a gallery images of people similar to a person query image. We learn good features for both MTMCT and Re ... Full text Cite

Single-Frame Indexing for 3D Hand Pose Estimation

Conference Proceedings of the IEEE International Conference on Computer Vision · February 11, 2016 Hand pose estimation from 3D sensor data matches a point cloud to a hand model, and has broad applications from gestural interfaces to scene understanding. We propose a novel scheme to index into a database of precomputed hand poses to initialize the match ... Full text Cite

Performance measures and a data set for multi-target, multi-camera tracking

Conference Lecture Notes in Computer Science Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics · January 1, 2016 To help accelerate progress in multi-target, multi-camera tracking systems, we present (i) a new pair of precision-recall measures of performance that treats errors of all types uniformly and emphasizes correct identification over sources of error; (ii) th ... Full text Cite

Distance minimization for reward learning from scored trajectories

Conference 30th Aaai Conference on Artificial Intelligence Aaai 2016 · January 1, 2016 Many planning methods rely on the use of an immediate reward function as a portable and succinct representation of desired behavior. Rewards are often inferred from demonstrated behavior that is assumed to be near-optimal. We examine a framework, Distance ... Cite

Tracking multiple people online and in real time

Conference Lecture Notes in Computer Science · January 1, 2015 We cast the problem of tracking several people as a graph partitioning problem that takes the form of an NP-hard binary integer program. We propose a tractable, approximate, online solution through the combination of a multi-stage cascade and a sliding tem ... Full text Cite

How to dispatch observers to track an evolving boundary

Conference 2007 First ACM/IEEE International Conference on Distributed Smart Cameras · 2007 Cite

Center-based approach for computerized analysis of non-uniform growth patterns in melanocytic lesions

Conference JOURNAL OF INVESTIGATIVE DERMATOLOGY · March 1, 2004 Link to item Cite

A learning method for automated polyp detection

Conference Lecture Notes in Computer Science Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics · January 1, 2001 Adenomatous polyps in the colon have a high probability of developing into subsequent colorectal carcinoma, the second leading cause of cancer deaths in United States. In this paper, we propose a new method for computer-aided diagnosis of polyps. Initial w ... Full text Cite

Using optical flow fields for Polyp detection in virtual colonoscopy

Conference Lecture Notes in Computer Science Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics · January 1, 2001 Since the introduction of Computed Tomographic Colonography (CTC), research has mainly focused on visualization and navigation techniques. Recently, efforts have shifted towards computer aided detection (CAD) of polyps. We propose a new approach to CAD in ... Full text Cite

Adaptive color-image embeddings for database navigation

Conference Lecture Notes in Computer Science · January 1, 1997 We present a novel approach to the problem of navigating through a database of color images for the purpose of image retrieval. We endow the database with a metric for the color distributions of the images. We then use multi-dimensional scaling techniques ... Full text Cite

Stereo without search

Conference Lecture Notes in Computer Science Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics · January 1, 1996 In its traditional formulation, stereo correspondence involves both searching and selecting. Given a feature in one scanline, the corresponding scanline in the other image is searched for the positions of similar features. Often more than one candidate is ... Full text Cite