paper-with-me

Papers

Multi-view Tracking Using Weakly Supervised Human Motion Prediction

2022-10-19 · Martin Engilberge, Weizhe Liu, Pascal Fua

Multi-view approaches to people-tracking have the potential to better handle occlusions than single-view ones in crowded scenes. They often rely on the tracking-by-detection paradigm, which involves detecting people first and then connecting the detections. In this paper, we argue that an even more effective approach is to predict people motion over time and infer people's presence in individual frames from these. This enables to enforce consistency both over time and across views of a single temporal frame. We validate our approach on the PETS2009 and WILDTRACK datasets and demonstrate that it outperforms state-of-the-art methods.

📄 PDF Abstract BibTeX arXiv:2210.10771

Code (1)

cvlab-epfl/mvflow 공식 구현 pytorch

Tasks

Human motion predictionmotion predictionMulti-Object TrackingMultiview DetectionPrediction

Similar Papers 제목 키워드 기반

Weakly-Supervised 3D Human Pose Learning via Multi-view Images in the Wild

2020-03-17 · CVPR 2020 6 · Umar Iqbal, Pavlo Molchanov, Jan Kautz

One major challenge for monocular 3D human pose estimation in-the-wild is the acquisition of training data that contains unconstrained images annotated with accurate 3D poses. In this paper, we address this challenge by …

3D Human Pose EstimationMonocular 3D Human Pose EstimationPose EstimationWeakly-superavised 3D Human Pose Estimation+1

Weakly Supervised Multi-Object Tracking and Segmentation

2021-01-03 · Idoia Ruiz, Lorenzo Porzi, Samuel Rota Bulò, Peter Kontschieder 외

We introduce the problem of weakly supervised Multi-Object Tracking and Segmentation, i.e. joint weakly supervised instance segmentation and multi-object tracking, in which we do not provide any kind of mask annotation. …

Instance SegmentationMulti-Object TrackingMulti-Object Tracking and SegmentationMulti-Task Learning+6

ViewSAM: Learning View-aware Cross-modal Semantics for Weakly Supervised Cross-view Referring Multi-Object Tracking

2026-05-04 · Jiawei Ge, Xintian Zhang, Jiuxin Cao, Bo Liu 외 arxiv

Cross-view Referring Multi-Object Tracking (CRMOT) aims to track multiple objects specified by natural language across multiple camera views, with globally consistent identities. Despite recent progress, existing methods…

Multi-Object Tracking

Weakly-supervised 3D Human Pose Estimation with Cross-view U-shaped Graph Convolutional Network

2021-05-23 · Guoliang Hua, Hong Liu, Wenhao Li, Qian Zhang 외

Although monocular 3D human pose estimation methods have made significant progress, it is far from being solved due to the inherent depth ambiguity. Instead, exploiting multi-view information is a practical way to achiev…

3D Human Pose EstimationMonocular 3D Human Pose EstimationPose EstimationWeakly-supervised 3D Human Pose Estimation+1

Video Object Segmentation and Tracking: A Survey

2019-04-19 · Rui Yao, Guosheng Lin, Shixiong Xia, Jiaqi Zhao 외

Object segmentation and object tracking are fundamental research area in the computer vision community. These two topics are diffcult to handle some common challenges, such as occlusion, deformation, motion blur, and sca…

Autonomous VehiclesObjectObject TrackingSegmentation+6