paper-with-me

Papers

CycAs: Self-supervised Cycle Association for Learning Re-identifiable Descriptions

2020-07-15 · ECCV 2020 8 · Zhongdao Wang, Jingwei Zhang, Liang Zheng, Yixuan Liu, Yifan Sun, Ya-Li Li, Shengjin Wang

This paper proposes a self-supervised learning method for the person re-identification (re-ID) problem, where existing unsupervised methods usually rely on pseudo labels, such as those from video tracklets or clustering. A potential drawback of using pseudo labels is that errors may accumulate and it is challenging to estimate the number of pseudo IDs. We introduce a different unsupervised method that allows us to learn pedestrian embeddings from raw videos, without resorting to pseudo labels. The goal is to construct a self-supervised pretext task that matches the person re-ID objective. Inspired by the \emph{data association} concept in multi-object tracking, we propose the \textbf{Cyc}le \textbf{As}sociation (\textbf{CycAs}) task: after performing data association between a pair of video frames forward and then backward, a pedestrian instance is supposed to be associated to itself. To fulfill this goal, the model must learn a meaningful representation that can well describe correspondences between instances in frame pairs. We adapt the discrete association process to a differentiable form, such that end-to-end training becomes feasible. Experiments are conducted in two aspects: We first compare our method with existing unsupervised re-ID methods on seven benchmarks and demonstrate CycAs' superiority. Then, to further validate the practical value of CycAs in real-world applications, we perform training on self-collected videos and report promising performance on standard test sets.

📄 PDF Abstract BibTeX arXiv:2007.07577

Code (0)

등록된 구현이 없습니다.

Tasks

ClusteringMulti-Object TrackingObject TrackingPerson Re-IdentificationSelf-Supervised Learning

Similar Papers 제목 키워드 기반

Generalizable Re-Identification from Videos with Cycle Association

2022-11-07 · Zhongdao Wang, Zhaopeng Dou, Jingwei Zhang, Liang Zheng 외

In this paper, we are interested in learning a generalizable person re-identification (re-ID) representation from unlabeled videos. Compared with 1) the popular unsupervised re-ID setting where the training and test sets…

Domain GeneralizationGeneralizable Person Re-identificationPerson Re-IdentificationSelf-Supervised Learning

Identifiable Deep Generative Models via Sparse Decoding

2021-10-20 · Gemma E. Moran, Dhanya Sridhar, Yixin Wang, David M. Blei

We develop the sparse VAE for unsupervised representation learning on high-dimensional data. The sparse VAE learns a set of latent factors (representations) which summarize the associations in the observed data features.…

Representation Learning

Identifying species network features from gene tree quartets under the coalescent model

2017-11-28

We show that many topological features of level-1 species networks are identifiable from the distribution of the gene tree quartets under the network multi-species coalescent model. In particular, every cycle of size at …

Image Understands Point Cloud: Weakly Supervised 3D Semantic Segmentation via Association Learning

2022-09-16 · Tianfang Sun, Zhizhong Zhang, Xin Tan, Yanyun Qu 외

Weakly supervised point cloud semantic segmentation methods that require 1\% or fewer labels, hoping to realize almost the same performance as fully supervised approaches, which recently, have attracted extensive researc…

3D Semantic SegmentationPseudo LabelSemantic SegmentationSuperpixels+1

Uncertainty-aware Self-supervised 3D Data Association

2020-08-18 · Jianren Wang, Siddharth Ancha, Yi-Ting Chen, David Held

3D object trackers usually require training on large amounts of annotated data that is expensive and time-consuming to collect. Instead, we propose leveraging vast unlabeled datasets by self-supervised metric learning of…

Metric LearningObjectobject-detectionObject Detection