Lune

ACM MM2022Top-tier venue

Revisiting Stochastic Learning for Generalizable Person Re-identification

Jiajian Zhao, Yifan Zhao, Xiaowu Chen, Jia Li

Abstract

Generalizable person re-identification aims to achieve a well generalization capability on target domains without accessing target data. Existing methods focus on suppressing domain-specific information or simulating unseen environments by meta-learning strategies, which could damage the capture ability on fine-grained visual patterns or lead to overfitting issues by the repetitive training of episodes. In this paper, we revisit the stochastic behaviors from two different perspectives: 1) Stochastic splitting-sliding sampler. It splits domain sources into approximately equal sample-size subsets and selects several subsets from various sources by a sliding window, forcing the model to step out of local minimums under stochastic sources. 2) Variance-varying gradient dropout. Gradients in parts of network are also selected by a sliding window and multiplied by binary masks generated from Bernoulli distribution, making gradients in varying variance and preventing the model from local minimums. By applying these two proposed stochastic behaviors, the model achieves a better generalization performance on unseen target domains without any additional computation costs or auxiliary modules. Extensive experiments demonstrate that our proposed model is effective and outperforms state-of-the-art methods on public domain generalizable person Re-ID benchmarks.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

Questions to start from

Lune can draw on

the full paper · related work

Get the answer

Free to start. No credit card required.

Lineage

Builds onCited byLeft to right by year; the more cited, the higher.

Builds on13

  1. Domain Generalization with MixStyleKaiyang Zhou, Yongxin Yang, Yu Qiao, Tao XiangICLR 2021 · 986 citations
  2. Domain Generalization using Causal MatchingDivyat Mahajan, Shruti Tople, Amit SharmaICML 2021 · 399 citations
  3. Just Pick a Sign: Optimizing Deep Multitask Models with Gradient Sign DropoutZhao Chen, Jiquan Ngiam, Yanping Huang, Thang Luong et al.NeurIPS 2020 · 313 citations
  4. Robust and Generalizable Visual Representation Learning via Random ConvolutionsZhenlin Xu, Deyi Liu, Junlin Yang, Colin Raffel et al.ICLR 2021 · 268 citations
  5. Self-Training With Progressive Augmentation for Unsupervised Cross-Domain Person Re-IdentificationXinyu Zhang, Jiewei Cao, Chunhua Shen, Mingyu YouICCV 2019 · 240 citations

Related papers

  1. Meta Batch-Instance Normalization for Generalizable Person Re-IdentificationSeokeon Choi, Taekyung Kim, Minki Jeong, Hyoungseob Park et al.CVPR 2021
  2. Style-Controllable Generalized Person Re-identificationYuke Li, Jingkuan Song, Hao Ni, Heng Tao ShenACM MM 2023 · 24 citations
  3. Meta Distribution Alignment for Generalizable Person Re-IdentificationHao Ni, Jingkuan Song, Xiaopeng Luo, Feng Zheng et al.CVPR 2022 · 77 citations
  4. Learning to Generalize Unseen Domains via Memory-based Multi-Source Meta-Learning for Person Re-IdentificationYuyang Zhao, Zhun Zhong, Fengxiang Yang, Zhiming Luo et al.CVPR 2021
  5. Debiased Batch Normalization via Gaussian Process for Generalizable Person Re-identificationJiawei Liu, Zhipeng Huang, Liang Li, Kecheng Zheng et al.AAAI 2022 · 30 citations
Dusk over the sea between two cliffs drawn in fine vertical lines