Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

Generalizable Person Re-identification via Balancing Alignment and Uniformity

T0 review · 3 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Data augmentations can hurt cross-domain person re-ID by shrinking feature diversity, and a new loss balance restores it.

desk verdict Solid empirical paper; the augmentation-polarization analysis is genuinely useful, but the causal story is under-supported and the headline numbers may be tuned on the target domain. read the letter →

arxiv 2411.11471 v1 pith:X36BGIWW submitted 2024-11-18 cs.CV

Yoonki Cho , Jaeyoon Kim , Woo Jae Kim , Junsik Jung , Sung-eui Yoon This is my paper · ORCID
classification cs.CV
keywords domaingeneralizablepersonre-identificationdataaugmentationalignmentuniformityrepresentationlearningopen-setretrievaldomain-invariantfeatures
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Domain-generalizable person re-identification aims to match people across cameras and datasets never seen during training. The paper finds that common data augmentations such as Random Erasing and RandAugment improve accuracy on the training domain but can degrade accuracy on unseen domains, a 'polarized effect.' The authors show that this effect coincides with a sparser feature space: augmented training pulls same-identity features closer (better alignment) but spreads the overall embedding distribution less evenly (worse uniformity). They propose BAU, which applies alignment and uniformity losses to both original and augmented images, down-weights unreliable augmented samples, and adds a domain-specific uniformity loss, and they report state-of-the-art results on standard benchmarks without complex architectures.

What carries the argument

The load-bearing object is the pair of loss terms borrowed from contrastive learning: alignment, the expected squared distance between positive pairs, and uniformity, the log of the average pairwise Gaussian potential on the unit hypersphere. BAU applies both to original and augmented images; the alignment loss is reweighted by the Jaccard similarity of k-reciprocal nearest-neighbor sets between augmented and original features, so that corrupted augmentations contribute less, and a third 'domain-specific uniformity' loss spreads each source domain's features against nearest class prototypes from the same domain in a memory bank. Together these terms are meant to keep the representation space both compact within identity and globally diverse, countering the sparsity that naive augmentation training induces.

What would settle it

Train a person re-ID model with aggressive augmentations at probability 1.0, then freeze the encoder and perform a post-hoc intervention that only raises uniformity, such as spectral decorrelation or a second training stage with a uniformity loss but no alignment term, and measure OOD mAP on Market-1501; if OOD accuracy does not improve even though uniformity is restored, the paper's claim that uniformity loss drives generalization is falsified. Alternatively, run MS+CS+C3 to Market-1501 with p=1.0 under three conditions, full BAU, BAU without the uniformity loss, and BAU with uniformity loss but alignment applied to randomly chosen pairs, and compare OOD mAP.

Watch

Extended reading notes

Core claim

The central discovery is that the polarized effect of data augmentations in DG re-ID is linked to a trade-off between alignment and uniformity on the hypersphere: augmentations improve alignment (invariance to transformations) but reduce uniformity (the even spread of embeddings), and because person re-ID is an open-set retrieval task, loss of uniformity means the model fails to preserve diverse visual cues needed for unseen identities and domains. To counter this, BAU explicitly optimizes both properties, using a reliability-weighted alignment loss between original and augmented views, a uniformity loss over both views, and a domain-specific uniformity loss computed with a momentum-updated memory bank of class prototypes. The paper reports that this simple regularization achieves state-of-the-art performance on Protocols 1 through 3, outperforming prior methods that rely on feature normalization, adversarial training, meta-learning, or mixture-of-experts.

Load-bearing premise

The argument assumes that the observed reduction in uniformity is the cause of the out-of-distribution performance drop, not merely a symptom of some other failure such as overfitting to spurious dominant features; if the causal arrow is wrong, BAU's gains could come from an unidentified mechanism.

Editorial extensions

If this is right

  • BAU turns data augmentations from a liability into a benefit: with the proposed losses, Random Erasing, RandAugment, and Color Jitter each improve out-of-distribution accuracy, whereas naive augmentation training degrades it.
  • The framework is architecture- and loss-agnostic: it improves ResNet-50, MobileNetV2, and ViT-B/16 baselines, and works on top of cross-entropy, triplet, ArcFace, and PCL losses.
  • Because BAU removes the need for complex components such as domain-adversarial training, meta-learning, or feature normalization, it offers a simpler recipe for DG re-ID that is easier to reproduce and extend.
  • If the analysis is right, the alignment-uniformity balance is the actionable quantity for diagnosing why an augmentation hurts generalization: a drop in uniformity flags the harmful regime.
  • The method's improvements hold across three evaluation protocols, including a leave-one-out setting where whole datasets are held out as unseen target domains.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Our inference: if uniformity loss is the causal mechanism, then any representation-learning task with an open-set or fine-grained retrieval component, such as vehicle re-ID or face recognition, should show the same polarized augmentation effect and benefit from a similar alignment-uniformity balance.
  • Our inference: the reliability weighting by k-reciprocal nearest-neighbor overlap is a general way to detect 'corrupted' augmentations and could be ported to other augmentation-heavy training regimes, such as semi-supervised or noisy-label learning.
  • Our inference: a direct test of the mechanism would be to add a uniformity regularizer to a model trained with aggressive augmentations but without the alignment weight; if out-of-distribution performance recovers only when both are present, the balance itself matters rather than uniformity alone.
  • Our inference: since Color Jitter alone did not show the polarized effect, a natural extension is to make augmentation selection adaptive, up-weighting transformations that preserve uniformity and down-weighting those that destroy it.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper studies the 'polarized effect' of data augmentation in domain-generalizable person re-identification (DG re-ID): augmentations improve in-distribution accuracy but degrade out-of-distribution accuracy. The authors attribute this to a reduction in feature-space uniformity, supported by correlational evidence in Fig. 1. They propose BAU, a training framework that adds alignment and uniformity losses on original and augmented images, a k-reciprocal-neighbor-based weighting of the alignment loss, and a domain-specific uniformity loss using class prototypes. The method is evaluated on three standard protocols across four datasets (Tables 3-4), with ablations in Tables 5-9 and qualitative analyses in Figs. 4-5. The empirical claim that BAU improves DG re-ID over prior state-of-the-art methods is well supported by the reported experiments.

Significance. If the mechanism is confirmed, the paper offers a simple, architecture-agnostic regularizer and a new explanation for the failure of data augmentation in open-set retrieval tasks. The manuscript's strengths include released code, comprehensive evaluation across multiple protocols and backbones/loss functions (Appendix B), component-wise ablations, and parameter sensitivity analyses. The headline accuracy gains are credible and sizable. However, the paper's scientific contribution hinges on the claim that augmentations hurt OOD performance specifically because they reduce uniformity; that causal link is currently underdetermined by the correlational and ablation evidence.

major comments (3)
  1. [Sec. 3.1 and Fig. 1] The central diagnostic claim—that data augmentations degrade OOD performance by reducing uniformity—is inferred from a correlation across five augmentation probabilities, with no error bars, no repeated seeds, and no interventional control. Increasing p also alters training difficulty, the degree of source overfitting, and the optimization geometry, all of which could independently produce both the mAP drop and the uniformity drop. As a result, the 'reveal' stated in Sec. 1 is not yet established. Please add an interventional experiment, e.g., train at a fixed high p while adding a uniformity-only regularizer, or compare with a matched non-uniformity regularizer (e.g., feature entropy maximization or feature-norm penalty) to show that restoring uniformity specifically recovers OOD mAP. If such an experiment is not feasible, the paper should reframe the finding as a correlational observation and present BAU as an empirically effective regularizer rather than a validated mechanism.
  2. [Table 5 and Eq. (9)] The ablation study supports Luniform as a useful component, but it never isolates Luniform alone among the BAU losses, nor does it compare Luniform against a generic regularizer of matched strength. The full objective also contains the weighting strategy and Ldomain (Eq. (9)), and Table 6 shows these components contribute. Therefore, the claim in Sec. 3.2 that the uniformity term is the key to generalization is underdetermined: the gains could come from extra regularization, from the weighting, or from the domain prototypes. Please include an ablation that adds only Luniform to the baseline (Lce+Ltri), and an additional control with a non-uniformity regularizer of similar complexity, so the mechanism can be separated from generic regularization.
  3. [Sec. 4.4 and Fig. 4b] The domain-specific uniformity loss (Eq. (8)) is claimed to 'reduce domain bias and thus enhance generalization,' but the t-SNE and uniformity measurements in Fig. 4b are computed on the source training data, not on the target domain. Since Ldomain contributes a large part of the overall gain (about 2.2 average mAP points in Table 6), the paper should provide target-domain evidence of domain invariance (e.g., uniformity of target features, or a domain-gap metric such as MMD between source and target feature distributions), or temper the claim to be about source-domain feature geometry only.
minor comments (5)
  1. [Tables 3-4 and Fig. 1] Report a single run without standard deviations; given the SOTA claims and the fact that Fig. 1 lacks error bars, the paper should report mean±std over at least three seeds for the main comparisons and the Fig. 1 analysis.
  2. [Eq. (4) and Sec. 4.2] The k-reciprocal Jaccard weight is computed within a mini-batch; the paper does not discuss sensitivity to batch size or to the sampling of identities (4 images per identity). Please note this as an approximation and add a sentence on how batch composition affects the reliability scores, or provide an experiment varying batch size.
  3. [Fig. 1c and Fig. 7] The y-axis label 'Uniformity (-Luniform)' is confusing because larger values indicate less uniformity; please relabel to '-Luniform (lower = more uniform)' or 'Inverse uniformity' to match the claims.
  4. [Sec. 3.1] The term 'sparse representation spaces' is used as a key explanation, but no formal definition or quantitative measure of sparsity is given (Grad-CAM in Fig. 2 is qualitative). Please define the sparsity metric or remove the term from the causal statement.
  5. [Eq. (7)-(8) and Fig. 8] Parameter analysis covers k and λ, but the momentum µ (Eq. 7) and the number of nearest prototypes N (Eq. 8) are set to fixed values without sensitivity analysis; please report a brief sensitivity check or justify these choices.

Circularity Check

1 steps flagged · score 1.0 of 10

No load-bearing circularity; BAU's SOTA results are empirical, though its uniformity-gap 'validation' is by construction because the loss explicitly optimizes that metric.

  1. self definitional [Sec. 4.4 'Analysis of alignment and uniformity', Fig. 4a; Eq. (9)]
    "In contrast, BAU consistently maintains better alignment and uniformity compared to the baseline across all augmentation probabilities. Notably, even when data augmentation is not applied (i.e., p = 0), BAU still outperforms the baseline."

    BAU's total objective in Eq. (9) explicitly contains Lalign, Luniform, and Ldomain, and Fig. 4a reports exactly these quantities. Therefore BAU's improved alignment/uniformity relative to the baseline is a direct consequence of optimizing those very terms, not an independent test of the claim that uniformity is the causal mediator of OOD generalization. This is a by-construction validation of a secondary diagnostic, not a prediction of the headline OOD accuracy.

full rationale

The paper's central assertion—that applying the BAU losses to original and augmented images improves out-of-distribution mAP—is an empirical result supported by Tables 3-5 and ablation studies; it is not algebraically derived from the loss definitions. No fitted parameter is renamed as a prediction, and no equation reduces to its own input. The causal diagnosis (augmentations reduce uniformity, and restoring uniformity is why BAU helps) rests on correlational evidence in Fig. 1b/c with no interventional control, but underdetermination of mechanism is a correctness/identification concern, not circularity. The one self-referential observation is Fig. 4a, where improved alignment/uniformity is guaranteed by Eq. (9) because those exact losses are being optimized; this weakens the mechanistic interpretation but does not invalidate the empirical SOTA claim. The only self-citation is Ref. [9] in the k-reciprocal weighting strategy, which is supported by four independent references and is ablated in Table 6, so it is not load-bearing. The Discussion's stated limitations about large domain shifts and the absence of adversarial/feature-level augmentations are acknowledged scope limits, not hidden circular inputs. Overall, no load-bearing circularity; score 1 reflects the minor by-construction uniformity validation.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The central claim rests on several domain assumptions about alignment and uniformity, a reliability-weighting hypothesis, and an implicitly tuned augmentation probability, plus three explicitly tuned hyperparameters. No new physical or conceptual entities are introduced. The main non-empirical risk is the causal chain from augmentation to reduced uniformity to OOD degradation, which is assumed rather than proven.

free parameters (6)
  • lambda_alignment_weight = 1.5
    Weights the alignment loss in Eq. 9; chosen for best mAP on MS+CS+C3 to Market-1501 in Fig 8b.
  • k_reciprocal_nearest_neighbors = 10
    Number of neighbors for the Jaccard weighting in Eq. 4; tuned on MS+CS+C3 to Market-1501 in Fig 8a.
  • augmentation_probability_p = 0.5
    Probability of applying augmentation during BAU training; selected as best in Table 9 on Protocol-3.
  • momentum_mu = 0.1
    Momentum coefficient for class prototype updates in Eq. 7; fixed implementation choice.
  • num_nearest_prototypes_N = mini-batch size
    Number of nearest same-domain prototypes used in Eq. 8; set to match the mini-batch size.
  • uniformity_and_domain_loss_weights = 1.0 each
    Coefficients for Luniform and Ldomain appear as fixed 1 in Eq. 9 and are not swept, which is a design choice affecting the balance.
assumptions (5)
  • domain assumption Uniformity of embeddings on the unit hypersphere is beneficial for generalization in open-set retrieval.
    Invoked in Sec 3.1 and 3.2 based on Wang and Isola and related work; the paper does not derive this property specifically for re-ID.
  • ad hoc to paper The k-reciprocal Jaccard similarity computed within a mini-batch measures whether an augmented sample is reliable for alignment learning.
    Used to weight Lalign in Eqs. 4 and 5; grounded in noisy-label literature but not independently validated for augmentation reliability in this setting.
  • ad hoc to paper Pushing features away from same-domain, different-class prototypes reduces domain bias and improves domain invariance.
    Defines Ldomain in Eq. 8; assumes that intra-domain uniformity yields domain-invariant features, which is a design hypothesis rather than a proven fact.
  • standard math Cross-entropy and batch-hard triplet losses provide a sound base objective for person re-identification.
    Adopted from the standard re-ID baseline in Eq. 9 and used as the backbone training objective.
  • domain assumption Mini-batch feature pairs approximate the global data distribution for computing uniformity.
    Luniform in Eq. 6 is computed over a mini-batch; the paper adds a memory bank because batches may not capture global structure, as stated in Sec 3.2.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Generalizable Person Re-identification via Balancing Alignment and Uniformity." pith.science (2026). https://pith.science/paper/X36BGIWW

@misc{pith2026241111471,
  author       = {Pith},
  title        = {Pith review of: Generalizable Person Re-identification via Balancing Alignment and Uniformity},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/X36BGIWW}},
  note         = {Machine review of arXiv:2411.11471}
}
read the original abstract

Domain generalizable person re-identification (DG re-ID) aims to learn discriminative representations that are robust to distributional shifts. While data augmentation is a straightforward solution to improve generalization, certain augmentations exhibit a polarized effect in this task, enhancing in-distribution performance while deteriorating out-of-distribution performance. In this paper, we investigate this phenomenon and reveal that it leads to sparse representation spaces with reduced uniformity. To address this issue, we propose a novel framework, Balancing Alignment and Uniformity (BAU), which effectively mitigates this effect by maintaining a balance between alignment and uniformity. Specifically, BAU incorporates alignment and uniformity losses applied to both original and augmented images and integrates a weighting strategy to assess the reliability of augmented samples, further improving the alignment loss. Additionally, we introduce a domain-specific uniformity loss that promotes uniformity within each source domain, thereby enhancing the learning of domain-invariant features. Extensive experimental results demonstrate that BAU effectively exploits the advantages of data augmentation, which previous studies could not fully utilize, and achieves state-of-the-art performance without requiring complex training procedures. The code is available at \url{https://github.com/yoonkicho/BAU}.

Figures

Figures reproduced from arXiv: 2411.11471 by the authors.

Figure 1
Figure 1. Analysis on polarized effect of data augmentations on in-distribution (ID) and out-of￾distribution (OOD). (a) mAP (%) on Market-1501 of models trained on the same dataset (ID) and MS+CS+C3 (OOD) with varying augmentation probabilities. (b) Alignment (Lalign) and uniformity (Luniform) of OOD scenarios (MS+CS+C3 → M). Counterintuitively, augmentations lead to more alignment but less uniformity, indicating that the mod… view at source ↗
Figure 2
Figure 2. Grad-CAM [68] across different probabili￾ties of data augmentations. To further explore the effect of data augmentations on the representa￾tion space, we leverage the concepts of alignment and uniformity [81], which are key properties of feature distributions on the unit hypersphere. Alignment is defined as the expected distance between positive pairs: Lalign ≜ log E (i,j)∼Ppos [∥fi − fj∥ 2 2 ], (1) where Ppos is th… view at source ↗
Figure 3
Figure 3. Overview of the proposed framework. In (b) and (c), each color represents a different identity and domain, respectively. (a) With original and augmented images, we apply alignment and uniformity losses to balance feature discriminability and generalization capability. We further introduce a domain-specific uniformity loss to mitigate domain bias. (b) Lalign pulls positive features closer, while Luniform pushes all f… view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Analysis of alignment and uniformity. (a) Alignment (Lalign) and uniformity (Luniform) on Market-1501 when MS+CS+C3 → M under Protocol-3 with varying augmentation probabilities. (b) T-SNE visualization with and without the domain-specific uniformity loss Ldomain. The v…
Figure 5
Figure 5. Figure 5: Analysis of the weighting strategy. (a) Quantitative comparison of mAP (%) across varying augmentation probabilities, with and without the weighting strategy, on MS+CS+C3 → M under Protocol-3. The weighting strategy consistently improves performance, especially at high…
Figure 6
Figure 6. Figure 6: Analysis on polarized effect across different types of (a) backbones, (b) loss functions, and (c) augmentations. The experimental configurations are the same in [PITH_FULL_IMAGE:figures/full_fig_p016_6.png]
Figure 7
Figure 7. Figure 7: Analysis on polarized effect of (a) Random Erasing and (b) RandAugment across augmentation probabilities. The experimental configurations are the same in [PITH_FULL_IMAGE:figures/full_fig_p016_7.png]
Figure 8
Figure 8. Figure 8: Parameter analysis of k and λ on MS+C3+CS → M under Protocol-3. (a) mAP/Rank-1 (%) with varying k-reciprocal nearest neighbors for the weighting strategy. (b) mAP/Rank-1 (%) with varying the weighting parameter λ for the alignment loss. Figure 8a shows the mAP and Rank…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Domain Generalization for Person Re-identification: A Survey Towards Domain-Agnostic Person Matching

    cs.CV 2025-06 conditional novelty 6.0 of 10

    A systematic survey that categorizes domain-generalizable person re-identification methods and compares their cross-domain performance.

Reference graph

Works this paper leans on

107 extracted references · 64 canonical work pages · cited by 1 Pith paper

  1. [1]

    An improved deep learning architecture for person re-identification

    Ejaz Ahmed, Michael Jones, and Tim K Marks. An improved deep learning architecture for person re-identification. In CVPR, 2015

    2015

  2. [2]

    Dex: Domain embedding expansion for generalized person re-identification

    Eugene PW Ang, Lin Shan, and Alex C Kot. Dex: Domain embedding expansion for generalized person re-identification. In BMVC, 2021

    2021

  3. [3]

    How to train your maml

    Antreas Antoniou, Harri Edwards, and Amos Storkey. How to train your maml. In ICLR, 2019

    2019

  4. [4]

    Person30k: A dual-meta generalization network for person re-identification

    Yan Bai, Jile Jiao, Wang Ce, Jun Liu, Yihang Lou, Xuetao Feng, and Ling-Yu Duan. Person30k: A dual-meta generalization network for person re-identification. In CVPR, 2021

    2021

  5. [5]

    Domain-specific batch normalization for unsupervised domain adaptation

    Woong-Gi Chang, Tackgeun You, Seonguk Seo, Suha Kwak, and Bohyung Han. Domain-specific batch normalization for unsupervised domain adaptation. In CVPR, 2019

    2019

  6. [6]

    Dual distribution alignment network for generalizable person re-identification

    Peixian Chen, Pingyang Dai, Jianzhuang Liu, Feng Zheng, Mingliang Xu, Qi Tian, and Rongrong Ji. Dual distribution alignment network for generalizable person re-identification. In AAAI, 2021

    2021

  7. [7]

    Abd-net: Attentive but diverse person re-identification

    Tianlong Chen, Shaojin Ding, Jingyi Xie, Ye Yuan, Wuyang Chen, Yang Yang, Zhou Ren, and Zhangyang Wang. Abd-net: Attentive but diverse person re-identification. In ICCV, 2019

    2019

  8. [8]

    Instance-dependent label-noise learning with manifold-regularized transition matrix estimation

    De Cheng, Tongliang Liu, Yixiong Ning, Nannan Wang, Bo Han, Gang Niu, Xinbo Gao, and Masashi Sugiyama. Instance-dependent label-noise learning with manifold-regularized transition matrix estimation. In CVPR, 2022

    2022

Show all 107 references
  1. [9]

    Part-based pseudo label refinement for unsupervised person re-identification

    Yoonki Cho, Woo Jae Kim, Seunghoon Hong, and Sung-Eui Yoon. Part-based pseudo label refinement for unsupervised person re-identification. In CVPR, 2022

    2022

  2. [10]

    Meta batch-instance normalization for generalizable person re-identification

    Seokeon Choi, Taekyung Kim, Minki Jeong, Hyoungseob Park, and Changick Kim. Meta batch-instance normalization for generalizable person re-identification. In CVPR, 2021

    2021

  3. [11]

    Randaugment: Practical automated data augmentation with a reduced search space

    Ekin D Cubuk, Barret Zoph, Jonathon Shlens, and Quoc V Le. Randaugment: Practical automated data augmentation with a reduced search space. In CVPR Workshops, 2020

    2020

  4. [12]

    Generalizable person re- identification with relevance-aware mixture of experts

    Yongxing Dai, Xiaotong Li, Jun Liu, Zekun Tong, and Ling-Yu Duan. Generalizable person re- identification with relevance-aware mixture of experts. In CVPR, 2021

    2021

  5. [13]

    Imagenet: A large-scale hierarchical image database

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. Imagenet: A large-scale hierarchical image database. In CVPR, 2009

    2009

  6. [14]

    Arcface: Additive angular margin loss for deep face recognition

    Jiankang Deng, Jia Guo, Niannan Xue, and Stefanos Zafeiriou. Arcface: Additive angular margin loss for deep face recognition. In CVPR, pages 4690–4699, 2019

    2019

  7. [15]

    Improved regularization of convolutional neural networks with cutout

    Terrance DeVries and Graham W Taylor. Improved regularization of convolutional neural networks with cutout. arXiv preprint arXiv:1708.04552, 2017

    2017 arXiv

  8. [16]

    An image is worth 16x16 words: Transformers for image recognition at scale

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, and Neil Houlsby. An image is worth 16x16 words: Transformers for image recognition at...

    2021

  9. [17]

    Identity-seeking self-supervised represen- tation learning for generalizable person re-identification

    Zhaopeng Dou, Zhongdao Wang, Yali Li, and Shengjin Wang. Identity-seeking self-supervised represen- tation learning for generalizable person re-identification. In ICCV, 2023

    2023

  10. [18]

    Model-agnostic meta-learning for fast adaptation of deep networks

    Chelsea Finn, Pieter Abbeel, and Sergey Levine. Model-agnostic meta-learning for fast adaptation of deep networks. In ICML, 2017

    2017

  11. [19]

    Self- similarity grouping: A simple unsupervised cross domain adaptation approach for person re-identification

    Yang Fu, Yunchao Wei, Guanshuo Wang, Yuqian Zhou, Honghui Shi, and Thomas S Huang. Self- similarity grouping: A simple unsupervised cross domain adaptation approach for person re-identification. In ICCV, 2019

    2019

  12. [20]

    Domain-adversarial training of neural networks.Journal of Machine Learning Research, 2016

    Yaroslav Ganin, Evgeniya Ustinova, Hana Ajakan, Pascal Germain, Hugo Larochelle, François Laviolette, Mario March, and Victor Lempitsky. Domain-adversarial training of neural networks.Journal of Machine Learning Research, 2016

    2016

  13. [21]

    Simcse: Simple contrastive learning of sentence embed- dings

    Tianyu Gao, Xingcheng Yao, and Danqi Chen. Simcse: Simple contrastive learning of sentence embed- dings. In EMNLP, 2021. 11

    2021

  14. [22]

    Mutual mean-teaching: Pseudo label refinery for unsuper- vised domain adaptation on person re-identification

    Yixiao Ge, Dapeng Chen, and Hongsheng Li. Mutual mean-teaching: Pseudo label refinery for unsuper- vised domain adaptation on person re-identification. In ICLR, 2020

    2020

  15. [23]

    Self-paced contrastive learning with hybrid memory for domain adaptive object re-id

    Yixiao Ge, Feng Zhu, Dapeng Chen, Rui Zhao, et al. Self-paced contrastive learning with hybrid memory for domain adaptive object re-id. NeurIPS, 2020

    2020

  16. [24]

    Imagenet-trained cnns are biased towards texture; increasing shape bias improves accuracy and robustness

    Robert Geirhos, Patricia Rubisch, Claudio Michaelis, Matthias Bethge, Felix A Wichmann, and Wieland Brendel. Imagenet-trained cnns are biased towards texture; increasing shape bias improves accuracy and robustness. In ICLR, 2019

    2019

  17. [25]

    Dissecting supervised contrastive learning

    Florian Graf, Christoph Hofer, Marc Niethammer, and Roland Kwitt. Dissecting supervised contrastive learning. In ICML, 2021

    2021

  18. [26]

    Viewpoint invariant pedestrian recognition with an ensemble of localized features

    Douglas Gray and Hai Tao. Viewpoint invariant pedestrian recognition with an ensemble of localized features. In ECCV, 2008

    2008

  19. [27]

    Generalizable person re-identification via self-supervised batch norm test-time adaption

    Ke Han, Chenyang Si, Yan Huang, Liang Wang, and Tieniu Tan. Generalizable person re-identification via self-supervised batch norm test-time adaption. In AAAI, 2022

    2022

  20. [28]

    Momentum contrast for unsupervised visual representation learning

    Kaiming He, Haoqi Fan, Yuxin Wu, Saining Xie, and Ross Girshick. Momentum contrast for unsupervised visual representation learning. In CVPR, 2020

    2020

  21. [29]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In CVPR, 2016

    2016

  22. [30]

    In defense of the triplet loss for person re- identification

    Alexander Hermans, Lucas Beyer, and Bastian Leibe. In defense of the triplet loss for person re- identification. arXiv preprint arXiv:1703.07737, 2017

    2017 arXiv

  23. [31]

    Person re-identification by descriptive and discriminative classification

    Martin Hirzer, Csaba Beleznai, Peter M Roth, and Horst Bischof. Person re-identification by descriptive and discriminative classification. In SCIA, 2011

    2011

  24. [32]

    Learning with neighbor consistency for noisy labels

    Ahmet Iscen, Jack Valmadre, Anurag Arnab, and Cordelia Schmid. Learning with neighbor consistency for noisy labels. In CVPR, 2022

    2022

  25. [33]

    Adaptive mixtures of local experts

    Robert A Jacobs, Michael I Jordan, Steven J Nowlan, and Geoffrey E Hinton. Adaptive mixtures of local experts. Neural Computation, 1991

    1991

  26. [34]

    Frustratingly easy person re-identification: Generaliz- ing person re-id in practice

    Jieru Jia, Qiuqi Ruan, and Timothy M Hospedales. Frustratingly easy person re-identification: Generaliz- ing person re-id in practice. In BMVC, 2019

    2019

  27. [35]

    Dynamically transformed instance normalization network for generalizable person re-identification

    Bingliang Jiao, Lingqiao Liu, Liying Gao, Guosheng Lin, Lu Yang, Shizhou Zhang, Peng Wang, and Yanning Zhang. Dynamically transformed instance normalization network for generalizable person re-identification. In ECCV, 2022

    2022

  28. [36]

    Style normalization and restitution for generalizable person re-identification

    Xin Jin, Cuiling Lan, Wenjun Zeng, Zhibo Chen, and Li Zhang. Style normalization and restitution for generalizable person re-identification. In CVPR, 2020

    2020

  29. [37]

    Cross-class feature augmentation for class incremental learning

    Taehoon Kim, Jaeyoo Park, and Bohyung Han. Cross-class feature augmentation for class incremental learning. In AAAI, 2024

    2024

  30. [38]

    Adam: A method for stochastic optimization

    Diederik P Kingma and Jimmy Ba. Adam: A method for stochastic optimization. In ICLR, 2015

    2015

  31. [39]

    Imagenet classification with deep convolutional neural networks

    Alex Krizhevsky, Ilya Sutskever, and Geoffrey E Hinton. Imagenet classification with deep convolutional neural networks. In NIPS, 2012

    2012

  32. [40]

    Learning to generalize: Meta-learning for domain generalization

    Da Li, Yongxin Yang, Yi-Zhe Song, and Timothy Hospedales. Learning to generalize: Meta-learning for domain generalization. In AAAI, 2018

    2018

  33. [41]

    Domain generalization with adversarial feature learning

    Haoliang Li, Sinno Jialin Pan, Shiqi Wang, and Alex C Kot. Domain generalization with adversarial feature learning. In CVPR, 2018

    2018

  34. [42]

    Prototypical contrastive learning of unsupervised representations

    Junnan Li, Pan Zhou, Caiming Xiong, and Steven Hoi. Prototypical contrastive learning of unsupervised representations. In ICLR, 2021

    2021

  35. [43]

    Locally aligned feature transforms across views

    Wei Li and Xiaogang Wang. Locally aligned feature transforms across views. In CVPR, 2013

    2013

  36. [44]

    Deepreid: Deep filter pairing neural network for person re-identification

    Wei Li, Rui Zhao, Tong Xiao, and Xiaogang Wang. Deepreid: Deep filter pairing neural network for person re-identification. In CVPR, 2014. 12

    2014

  37. [45]

    Harmonious attention network for person re-identification

    Wei Li, Xiatian Zhu, and Shaogang Gong. Harmonious attention network for person re-identification. In CVPR, 2018

    2018

  38. [46]

    Uncertainty modeling for out-of-distribution generalization

    Xiaotong Li, Yongxing Dai, Yixiao Ge, Jun Liu, Ying Shan, and Ling-Yu Duan. Uncertainty modeling for out-of-distribution generalization. In ICLR, 2022

    2022

  39. [47]

    Revisiting batch normalization for practical domain adaptation

    Yanghao Li, Naiyan Wang, Jianping Shi, Jiaying Liu, and Xiaodi Hou. Revisiting batch normalization for practical domain adaptation. arXiv preprint arXiv:1603.04779, 2016

    2016 arXiv

  40. [48]

    Style-controllable generalized person re- identification

    Yuke Li, Jingkuan Song, Hao Ni, and Heng Tao Shen. Style-controllable generalized person re- identification. In ACM MM, 2023

    2023

  41. [49]

    Mind the gap: Understanding the modality gap in multi-modal contrastive representation learning

    Victor Weixin Liang, Yuhui Zhang, Yongchan Kwon, Serena Yeung, and James Y Zou. Mind the gap: Understanding the modality gap in multi-modal contrastive representation learning. NeurIPS, 2022

    2022

  42. [50]

    Interpretable and generalizable person re-identification with query-adaptive convolution and temporal lifting

    Shengcai Liao and Ling Shao. Interpretable and generalizable person re-identification with query-adaptive convolution and temporal lifting. In ECCV, 2020

    2020

  43. [51]

    Transmatcher: Deep image matching through transformers for generalizable person re-identification

    Shengcai Liao and Ling Shao. Transmatcher: Deep image matching through transformers for generalizable person re-identification. In NeurIPS, 2021

    2021

  44. [52]

    Graph sampling based deep metric learning for generalizable person re-identification

    Shengcai Liao and Ling Shao. Graph sampling based deep metric learning for generalizable person re-identification. In CVPR, 2022

    2022

  45. [53]

    Debiased batch normalization via gaussian process for generalizable person re-identification

    Jiawei Liu, Zhipeng Huang, Liang Li, Kecheng Zheng, and Zheng-Jun Zha. Debiased batch normalization via gaussian process for generalizable person re-identification. In AAAI, 2022

    2022

  46. [54]

    Adaptive transfer network for cross-domain person re-identification

    Jiawei Liu, Zheng-Jun Zha, Di Chen, Richang Hong, and Meng Wang. Adaptive transfer network for cross-domain person re-identification. In CVPR, 2019

    2019

  47. [55]

    Multi-camera activity correlation analysis

    Chen Change Loy, Tao Xiang, and Shaogang Gong. Multi-camera activity correlation analysis. In CVPR, 2009

    2009

  48. [56]

    A strong baseline and batch normalization neck for deep person re-identification

    Hao Luo, Wei Jiang, Youzhi Gu, Fuxu Liu, Xingyu Liao, Shenqi Lai, and Jianyang Gu. A strong baseline and batch normalization neck for deep person re-identification. IEEE Transactions on Multimedia, 2019

    2019

  49. [57]

    Coco-lm: Correcting and contrasting text sequences for language model pretraining

    Yu Meng, Chenyan Xiong, Payal Bajaj, Paul Bennett, Jiawei Han, Xia Song, et al. Coco-lm: Correcting and contrasting text sequences for language model pretraining. NeurIPS, 2021

    2021

  50. [58]

    Diva: Diverse visual feature aggregation for deep metric learning

    Timo Milbich, Karsten Roth, Homanga Bharadhwaj, Samarth Sinha, Yoshua Bengio, Björn Ommer, and Joseph Paul Cohen. Diva: Diverse visual feature aggregation for deep metric learning. In ECCV, 2020

    2020

  51. [59]

    Learning from failure: De- biasing classifier from biased classifier

    Junhyun Nam, Hyuntak Cha, Sungsoo Ahn, Jaeho Lee, and Jinwoo Shin. Learning from failure: De- biasing classifier from biased classifier. NeurIPS, 2020

    2020

  52. [60]

    Part-aware transformer for generalizable person re-identification

    Hao Ni, Yuke Li, Lianli Gao, Heng Tao Shen, and Jingkuan Song. Part-aware transformer for generalizable person re-identification. In ICCV, 2023

    2023

  53. [61]

    Meta distribution alignment for generalizable person re-identification

    Hao Ni, Jingkuan Song, Xiaopeng Luo, Feng Zheng, Wen Li, and Heng Tao Shen. Meta distribution alignment for generalizable person re-identification. In CVPR, 2022

    2022

  54. [62]

    On first-order meta-learning algorithms

    A Nichol. On first-order meta-learning algorithms. arXiv preprint arXiv:1803.02999, 2018

    2018 arXiv

  55. [63]

    Geodesic multi-modal mixup for robust fine-tuning

    Changdae Oh, Junhyuk So, Hoyoon Byun, YongTaek Lim, Minchul Shin, Jong-June Jeon, and Kyungwoo Song. Geodesic multi-modal mixup for robust fine-tuning. NeurIPS, 2023

    2023

  56. [64]

    Pytorch: An imperative style, high- performance deep learning library

    Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, Zeming Lin, Natalia Gimelshein, Luca Antiga, et al. Pytorch: An imperative style, high- performance deep learning library. In NeurIPS, 2019

    2019

  57. [65]

    Alignment-uniformity aware representation learning for zero-shot video classification

    Shi Pu, Kaili Zhao, and Mao Zheng. Alignment-uniformity aware representation learning for zero-shot video classification. In CVPR, 2022

    2022

  58. [66]

    A closer look at smoothness in domain adversarial training

    Harsh Rangwani, Sumukh K Aithal, Mayank Mishra, Arihant Jain, and Venkatesh Babu Radhakrishnan. A closer look at smoothness in domain adversarial training. In ICML, 2022

    2022

  59. [67]

    Mobilenetv2: Inverted residuals and linear bottlenecks

    Mark Sandler, Andrew Howard, Menglong Zhu, Andrey Zhmoginov, and Liang-Chieh Chen. Mobilenetv2: Inverted residuals and linear bottlenecks. In CVPR, 2018. 13

    2018

  60. [68]

    Grad-cam: Visual explanations from deep networks via gradient-based localization

    Ramprasaath R Selvaraju, Michael Cogswell, Abhishek Das, Ramakrishna Vedantam, Devi Parikh, and Dhruv Batra. Grad-cam: Visual explanations from deep networks via gradient-based localization. In ICCV, 2017

    2017

  61. [69]

    Generalizing across domains via cross-gradient training

    Shiv Shankar, Vihari Piratla, Soumen Chakrabarti, Siddhartha Chaudhuri, Preethi Jyothi, and Sunita Sarawagi. Generalizing across domains via cross-gradient training. In ICLR, 2018

    2018

  62. [70]

    Very deep convolutional networks for large-scale image recognition

    Karen Simonyan and Andrew Zisserman. Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556, 2014

    2014 arXiv

  63. [71]

    Generalizable person re-identification by domain-invariant mapping network

    Jifei Song, Yongxin Yang, Yi-Zhe Song, Tao Xiang, and Timothy M Hospedales. Generalizable person re-identification by domain-invariant mapping network. In CVPR, 2019

    2019

  64. [72]

    Pose-driven deep convolutional model for person re-identification

    Chi Su, Jianing Li, Shiliang Zhang, Junliang Xing, Wen Gao, and Qi Tian. Pose-driven deep convolutional model for person re-identification. In ICCV, 2017

    2017

  65. [73]

    Beyond part models: Person retrieval with refined part pooling (and a strong convolutional baseline)

    Yifan Sun, Liang Zheng, Yi Yang, Qi Tian, and Shengjin Wang. Beyond part models: Person retrieval with refined part pooling (and a strong convolutional baseline). In ECCV, 2018

    2018

  66. [74]

    Visualizing data using t-sne

    Laurens Van der Maaten and Geoffrey Hinton. Visualizing data using t-sne. Journal of Machine Learning Research, 2008

    2008

  67. [75]

    Generalizing to unseen domains via adversarial data augmentation

    Riccardo V olpi, Hongseok Namkoong, Ozan Sener, John C Duchi, Vittorio Murino, and Silvio Savarese. Generalizing to unseen domains via adversarial data augmentation. In NeurIPS, 2018

    2018

  68. [76]

    Tent: Fully test-time adaptation by entropy minimization

    Dequan Wang, Evan Shelhamer, Shaoteng Liu, Bruno Olshausen, and Trevor Darrell. Tent: Fully test-time adaptation by entropy minimization. In ICLR, 2021

    2021

  69. [77]

    Understanding the behaviour of contrastive loss

    Feng Wang and Huaping Liu. Understanding the behaviour of contrastive loss. In CVPR, 2021

    2021

  70. [78]

    Learning discriminative features with multiple granularities for person re-identification

    Guanshuo Wang, Yufeng Yuan, Xiong Chen, Jiwei Li, and Xi Zhou. Learning discriminative features with multiple granularities for person re-identification. In ACM MM, 2018

    2018

  71. [79]

    Generalizing to unseen domains: A survey on domain generalization

    Jindong Wang, Cuiling Lan, Chang Liu, Yidong Ouyang, Tao Qin, Wang Lu, Yiqiang Chen, Wenjun Zeng, and S Yu Philip. Generalizing to unseen domains: A survey on domain generalization. IEEE Transactions on Knowledge and Data Engineering , 2022

    2022

  72. [80]

    Feature alignment and uniformity for test time adaptation

    Shuai Wang, Daoan Zhang, Zipei Yan, Jianguo Zhang, and Rui Li. Feature alignment and uniformity for test time adaptation. In CVPR, 2023

    2023

  73. [81]

    Understanding contrastive representation learning through alignment and uniformity on the hypersphere

    Tongzhou Wang and Phillip Isola. Understanding contrastive representation learning through alignment and uniformity on the hypersphere. In ICML, 2020

    2020

  74. [82]

    Domainmix: Learning generalizable person re-identification without human annotations

    Wenhao Wang, Shengcai Liao, Fang Zhao, Cuicui Kang, and Ling Shao. Domainmix: Learning generalizable person re-identification without human annotations. In BMVC, 2021

    2021

  75. [83]

    Person transfer gan to bridge domain gap for person re-identification

    Longhui Wei, Shiliang Zhang, Wen Gao, and Qi Tian. Person transfer gan to bridge domain gap for person re-identification. In CVPR, 2018

    2018

  76. [84]

    A topological filter for learning with label noise

    Pengxiang Wu, Songzhu Zheng, Mayank Goswami, Dimitris Metaxas, and Chao Chen. A topological filter for learning with label noise. NeurIPS, 2020

    2020

  77. [85]

    Joint detection and identification feature learning for person search

    Tong Xiao, Shuang Li, Bochao Wang, Liang Lin, and Xiaogang Wang. Joint detection and identification feature learning for person search. In CVPR, 2017

    2017

  78. [86]

    Mimic embedding via adaptive aggregation: learning generalizable person re-identification

    Boqiang Xu, Jian Liang, Lingxiao He, and Zhenan Sun. Mimic embedding via adaptive aggregation: learning generalizable person re-identification. In ECCV, 2022

    2022

  79. [87]

    Learning with diversity: Self- expanded equalization for better generalized deep metric learning

    Jiexi Yan, Zhihui Yin, Erkun Yang, Yanhua Yang, and Heng Huang. Learning with diversity: Self- expanded equalization for better generalized deep metric learning. In ICCV, 2023

    2023

  80. [88]

    Deep learning for person re-identification: A survey and outlook

    Mang Ye, Jianbing Shen, Gaojie Lin, Tao Xiang, Ling Shao, and Steven CH Hoi. Deep learning for person re-identification: A survey and outlook. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021

    2021

  81. [89]

    Cutmix: Regularization strategy to train strong classifiers with localizable features

    Sangdoo Yun, Dongyoon Han, Seong Joon Oh, Sanghyuk Chun, Junsuk Choe, and Youngjoon Yoo. Cutmix: Regularization strategy to train strong classifiers with localizable features. In ICCV, 2019

    2019

  82. [90]

    Adaptive cross-domain learning for generalizable person re-identification

    Pengyi Zhang, Huanzhang Dou, Yunlong Yu, and Xi Li. Adaptive cross-domain learning for generalizable person re-identification. In ECCV, 2022. 14

    2022

  83. [91]

    Learning domain invariant representations for generalizable person re-identification

    Yi-Fan Zhang, Zhang Zhang, Da Li, Zhen Jia, Liang Wang, and Tieniu Tan. Learning domain invariant representations for generalizable person re-identification. IEEE Transactions on Image Processing, 2022

    2022

  84. [92]

    Generalized cross entropy loss for training deep neural networks with noisy labels

    Zhilu Zhang and Mert Sabuncu. Generalized cross entropy loss for training deep neural networks with noisy labels. NeurIPS, 2018

    2018

  85. [93]

    Maximum-entropy adversarial data augmentation for improved generalization and robustness

    Long Zhao, Ting Liu, Xi Peng, and Dimitris Metaxas. Maximum-entropy adversarial data augmentation for improved generalization and robustness. In NeruIPS, 2020

    2020

  86. [94]

    Learning to generalize unseen domains via memory-based multi-source meta-learning for person re- identification

    Yuyang Zhao, Zhun Zhong, Fengxiang Yang, Zhiming Luo, Yaojin Lin, Shaozi Li, and Nicu Sebe. Learning to generalize unseen domains via memory-based multi-source meta-learning for person re- identification. In CVPR, 2021

    2021

  87. [95]

    Scalable person re-identification: A benchmark

    Liang Zheng, Liyue Shen, Lu Tian, Shengjin Wang, Jingdong Wang, and Qi Tian. Scalable person re-identification: A benchmark. In ICCV, 2015

    2015

  88. [96]

    Person re-identification: Past, present and future

    Liang Zheng, Yi Yang, and Alexander G Hauptmann. Person re-identification: Past, present and future. arXiv preprint arXiv:1610.02984, 2016

    2016 arXiv

  89. [97]

    Associating groups of people

    Wei-Shi Zheng, Shaogang Gong, and Tao Xiang. Associating groups of people. In BMVC, 2009

    2009

  90. [98]

    Unlabeled samples generated by gan improve the person re-identification baseline in vitro

    Zhedong Zheng, Liang Zheng, and Yi Yang. Unlabeled samples generated by gan improve the person re-identification baseline in vitro. In ICCV, 2017

    2017

  91. [99]

    Adversarial style augmentation for domain generalized urban-scene segmentation

    Zhun Zhong, Yuyang Zhao, Gim Hee Lee, and Nicu Sebe. Adversarial style augmentation for domain generalized urban-scene segmentation. In NeurIPS, 2022

    2022

  92. [100]

    Re-ranking person re-identification with k-reciprocal encoding

    Zhun Zhong, Liang Zheng, Donglin Cao, and Shaozi Li. Re-ranking person re-identification with k-reciprocal encoding. In CVPR, 2017

    2017

  93. [101]

    Random erasing data augmentation

    Zhun Zhong, Liang Zheng, Guoliang Kang, Shaozi Li, and Yi Yang. Random erasing data augmentation. In AAAI, 2020

    2020

  94. [102]

    Invariance matters: Exemplar memory for domain adaptive person re-identification

    Zhun Zhong, Liang Zheng, Zhiming Luo, Shaozi Li, and Yi Yang. Invariance matters: Exemplar memory for domain adaptive person re-identification. In CVPR, 2019

    2019

  95. [103]

    Domain generalization: A survey

    Kaiyang Zhou, Ziwei Liu, Yu Qiao, Tao Xiang, and Chen Change Loy. Domain generalization: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence , 2022

    2022

  96. [104]

    Deep domain-adversarial image generation for domain generalisation

    Kaiyang Zhou, Yongxin Yang, Timothy Hospedales, and Tao Xiang. Deep domain-adversarial image generation for domain generalisation. In AAAI, 2020

    2020

  97. [105]

    Learning to generate novel domains for domain generalization

    Kaiyang Zhou, Yongxin Yang, Timothy Hospedales, and Tao Xiang. Learning to generate novel domains for domain generalization. In ECCV, 2020

    2020

  98. [106]

    Domain generalization with mixstyle

    Kaiyang Zhou, Yongxin Yang, Yu Qiao, and Tao Xiang. Domain generalization with mixstyle. In ICLR, 2021

    2021

  99. [107]

    Rethinking the distribution gap of person re-identification with camera-based batch normalization

    Zijie Zhuang, Longhui Wei, Lingxi Xie, Tianyu Zhang, Hengheng Zhang, Haozhe Wu, Haizhou Ai, and Qi Tian. Rethinking the distribution gap of person re-identification with camera-based batch normalization. In ECCV, 2020. 15 A Additional Implementation Details This section provid...

    2020

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.