Pith. sign in

REVIEW 4 major objections 5 minor 71 references

Semantic-Aligned Learning with Collaborative Refinement for Unsupervised VI-ReID

T0 review · 4 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read This paper claims that unsupervised matching of people across visible and infrared cameras improves when fine-grained part features are aligned in each modality's own pseudo-label space, rather than relying on global features alone.

desk verdict A thoroughly ablated USL-VI-ReID system whose most defensible contribution is a portable feature-level re-ranking; the fine-grained alignment module is plausible but its key cross-modality correspondence assumption is unvalidated. read the letter →

arxiv 2504.19244 v2 pith:DOWFKYLB submitted 2025-04-27 cs.CV

De Cheng , Lingfeng He , Nannan Wang , Dingwen Zhang , Xinbo Gao This is my paper · ORCID
classification cs.CV
keywords unsupervisedvisible-infraredpersonre-identificationcross-modalitymatchingfine-grainedsemanticalignmentquery-guidedattentionpseudo-labelrefinementcontrastivelearningfeaturepropagationre-ranking
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

Unsupervised visible-infrared person re-identification aims to match the same person across daylight and thermal camera images with no identity labels during training. The paper claims that existing methods fall short because they only align global, whole-image features, leaving fine-grained patterns — color details that visible images emphasize and contour and edge cues that infrared images emphasize — uncoordinated across modalities. To fix this, it proposes a framework that derives a query from the fine-grained patterns one modality cares about and uses that query to pull matching part-level patterns out of the paired image from the other modality, then trains those aligned parts against the corresponding pseudo-labels. It also adds an online refinement step that finds reliable positive samples to counter noisy pseudo-labels, and a feature-propagation re-ranking step that is fast because it works directly on features rather than on a final distance matrix. If the claims hold, this is a step toward practical surveillance systems that can search for a person across visible and infrared cameras without manual annotation effort.

What carries the argument

The load-bearing object is the instance-adaptive query: for each of $N_p$ body parts, the pixel features in that part’s sequence are weighted by their similarity to a learnable modality-shared part prototype and summed into a query vector. This query is then used as the Q in scaled dot-product attention over the same part region in a paired cross-modality image, so the resulting feature aggregates exactly the pixels that share the query’s semantics. The DAGL module supplies the pairs by using an optimal-transport label assignment to give every infrared image a visible-cluster pseudo-label and vice versa, while the GPCR module refines those noisy label structures with online positive mining; the CMFP module enriches features with neighborhood information through matrix operations on a symmetrized affinity graph. Together these components convert global-only cross-modality contrastive learning into part-level contrastive learning driven by two complementary label spaces.

What would settle it

Take the trained SALCR setting, randomly permute between 10% and 50% of the cross-modality pseudo-label pairs that define FGSAL’s potential positive pairs, retrain from scratch, and track Rank-1 and mAP; if the fine-grained part-level loss adds little once even 20% of the pairs are wrong, the method’s gain is hostage to the dual association’s accuracy, whereas if performance degrades gracefully past 50%, GPCR’s refinement is doing most of the work.

Watch

Extended reading notes

Core claim

At the center of the paper is the observation that pseudo-label distributions of visible and infrared clusters diverge: some clusters contain almost only one modality, reflecting fine-grained features that are distinctive in one modality but weak in the other. The proposed SALCR exploits this by maintaining two label spaces, one per modality, and pushing cross-modality part features to agree in whichever label space emphasizes them. The FGSAL module creates an instance-adaptive query per body part by pooling within a part region, weighted by similarity to a learned part prototype; that query then acts as the attention key over the corresponding part region in a paired cross-modality image, producing semantic-aligned part features from both modalities. These features are optimized with part-level contrastive losses in the visible and infrared pseudo-label spaces. The GPCR module mines reliable positive sets online — a cross-modality intersection for global features and a mutual-correction strategy for part features — and the CMFP module propagates features along a k-nearest-neighbor affinity graph during both association and retrieval. The paper reports that the full framework outperforms the previous best unsupervised method GUR* by 3.45% mAP and 3.49% Rank-1 on SYSU-MM01 All-Search, and by 13.64% mAP and 16.67% Rank-1 on RegDB Visible-to-Infrared; with the test-time CMFP re-ranking it reaches 78.29% Rank-1 and 74.08% mAP on SYSU-MM01.

Load-bearing premise

The part-level alignment works only if the pseudo-label pairs that FGSAL treats as potential positives are mostly genuine same-identity pairs, because the query-guided attention has no independent check that the cross-modality image it is aligning to is the same person.

Editorial extensions

If this is right

  • On the two standard benchmarks, the framework reports state-of-the-art unsupervised results without test-time re-ranking: 64.44% Rank-1 and 60.44% mAP on SYSU-MM01 All-Search, and 90.58% Rank-1 and 83.87% mAP on RegDB Visible-to-Infrared, beating the previous best GUR*.
  • With CMFP applied as a test-time re-ranking, retrieval improves to 78.29% Rank-1 and 74.08% mAP on SYSU-MM01 and to 93.01% Rank-1 and 93.16% mAP on RegDB, while taking about 0.24 and 0.49 seconds respectively — much faster than the compared CMRR re-ranking.
  • Pseudo-label quality, measured by Adjusted Rand Index, Fowlkes-Mallows Index, Adjusted Mutual Information, and V-measure, rises over training when FGSAL and GPCR are added, so the learned cross-modality associations become increasingly identity-consistent.
  • CMFP can be attached to other unsupervised VI-ReID methods as a post-processing step; applied to PGM, MULT, and SDCL in the paper, it raises their Rank-1 and mAP on SYSU-MM01 by roughly 9 to 13 percentage points.
  • The extra part-feature stream adds only 0.35M parameters and 0.01 GMac of computation, so the fine-grained alignment cost is small relative to the reported mAP gain over global-only training.

Reading between the lines

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

  • A stress test the authors do not run but that would map this method’s limits is to corrupt a controlled fraction of the dual-association pseudo-label pairs and measure when FGSAL’s gain collapses; this would quantify how much association accuracy the semantic-alignment step requires.
  • Because CMFP is pure matrix multiplication on an affinity graph with a single tuning parameter, it should transfer to other cross-modal retrieval settings, such as text-image or sketch-image matching; the paper demonstrates it on three VI-ReID methods but not on other modalities.
  • The query-guided attention mechanism suggests visible-specific cues and infrared-specific cues could be learned separately and then fused; one could test whether explicitly decomposing color cues from shape and edge cues improves performance further than letting the two pseudo-label spaces do that work implicitly.
  • The part prototypes are learned without part annotations, so the same machinery could plausibly be adapted to unsupervised person re-identification across visible cameras with very different resolutions or viewpoints, where the per-modality label spaces would correspond to per-domain label spaces.
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

4 major / 5 minor

Summary. The paper proposes SALCR, an unsupervised visible-infrared person re-identification framework. Pseudo-labels are first generated intra-modality by DBSCAN and then unified across modalities by a dual optimal-transport association (DAGL). A Fine-Grained Semantic-Aligned Learning module (FGSAL) generates instance-adaptive queries from spatial part sequences and applies them to cross-modality feature maps to produce 'semantic-aligned' part features, which are optimized by part-level contrastive losses in the visible and infrared label spaces. A Global-Part Collaborative Refinement module (GPCR) mines neighbor-based positive sets for global and part features, and a Cross-Modality Feature Propagation module (CMFP) propagates features along k-NN graphs during training and testing. Experiments on SYSU-MM01 and RegDB report state-of-the-art results, together with ablations, hyperparameter analyses, pseudo-label quality metrics, attention visualizations, and a unified training-set experiment.

Significance. If the mechanism performs as claimed, the paper is a credible advance for USL-VI-ReID: it is among the first methods to use fine-grained, query-guided cross-modality part features, and the reported gains over prior unsupervised methods are substantial (e.g., 3.45% mAP and 3.49% Rank-1 over GUR* on SYSU-MM01 All-Search, and 13.64% mAP and 16.67% Rank-1 on RegDB V-to-I). The paper's strengths include controlled ablations for each module, per-hyperparameter analyses, pseudo-label quality evaluation across epochs, a unified training-set generalization experiment, a computational complexity analysis of CMFP, and a commitment to release code. The main weaknesses are the lack of direct validation that the FGSAL spatial-part correspondences are semantically aligned, internal notation inconsistencies in the FGSAL/GPCR feature definitions, and the absence of error bars or significance information for the small margins in the ablation and hyperparameter tables. These issues concern the central claim and need to be resolved before the results can be fully interpreted.

major comments (4)
  1. [Sec. 3.3, Eqs. (9)-(11)] The central assumption of FGSAL is that the p-th spatial block in the visible and infrared feature maps, combined with the query-guided attention, yields part features with the same identity-discriminative semantics. This assumption is not validated. On SYSU-MM01 the visible and infrared cameras have different viewpoints and the images are not pixel-aligned, so the contiguous spatial split used in Eq. (9) does not by itself guarantee semantic correspondence; the attention in Eqs. (10)-(11) only aggregates within the same block index. The paper acknowledges noisy pseudo-labels but never quantifies the fraction of wrong positive pairs or the rate of correct part correspondence. Consequently, the roughly 3 mAP gain attributed to FGSAL in Table 3 could come from the additional part-level contrastive loss rather than from semantic alignment. I would like to see a quantitative test, such as corrupting a fraction of the potential positive pairs, replacing the queries with random or fixed block pooling, and reporting the effect, together with a measure of part correspondence on the two datasets.
  2. [Sec. 3.3, Eqs. (10)-(11) and Sec. 3.4, Eq. (23)] The notation for the query-generated part features is internally inconsistent. Eq. (11) defines fvr_p = Avr_p * Fr_p^T, i.e., a visible query applied to the infrared sequence, but the following sentence says the pair {fvv_p, frv_p} is derived from qv_p, and Sec. 3.4, Eq. (23) lists {fvv_p, fav_p, frv_p} as a triplet from qv_p. As written, frv_p is later supervised in the visible label space in Eqs. (14) and (24)-(25), which only makes sense if frv_p denotes the infrared feature attended by the visible query. The definitions need to be aligned so that each part feature is unambiguously associated with its generating query and label space; currently the supervision of the FGSAL losses is not well-defined.
  3. [Sec. 3.4, Eqs. (16)-(17) and (21)-(23)] The positive sets in GPCR are defined as intersections of k-NN sets, for example Eq. (16) intersects N{fv, Mv_I, k} and N{fr, Mv_I, k}. There is no discussion of the case where this intersection is empty, which can occur especially in the early training epochs; in that case the contrastive loss in Eq. (18) has an empty numerator and is undefined. The authors should state the implementation fallback, such as discarding such pairs, using a union, or enforcing a minimum-size constraint, and report how often this situation occurs. This is necessary for the GPCR losses to be well-defined.
  4. [Tables 3 and 9-14] The component contributions are reported without error bars or significance tests, although several margins are small. For example, in Table 3 FGSAL adds 3.34 mAP on SYSU-MM01 All-Search, but in Table 9 the difference between Np=3 and Np=4 is only 0.88 mAP and Np=2 is actually better on Indoor-Search; in Tables 11-14 neighboring ktr/kte settings differ by less than 1 mAP on several settings. Since Np, k, ktr, and kte are tuned separately per dataset and the RegDB protocol already averages over 10 random splits, the authors should report standard deviations or a small number of repeated runs for the main tables and ablations so that the relative contributions of FGSAL and GPCR are distinguishable from training noise.
minor comments (5)
  1. [Table 1] The RegDB block repeats the header 'Visible-to-Infrared' for both protocol columns; the second column should be 'Infrared-to-Visible'.
  2. [Abstract and Sec. 3.2] The acronym for the dual association module is given as 'DAGI' in the abstract and as 'DAGL' in the method section; please unify the naming.
  3. [Sec. 4.5] The text refers to 'Fig.4.4' multiple times (e.g., 'too large lambda (lambda=2.0 in Fig.4.4)'); the figure references need to be corrected to the actual figure numbers.
  4. [Introduction and Sec. 3.5] The claim that CMFP is 'a more efficient and effective solution compared to existing cross-modality re-ranking technologies' is too strong given Table 6: on RegDB, CMRR with k=8 reaches 95.39/95.72 R1/mAP whereas CMFP(te) reaches 93.01/93.16. Please qualify the claim to specify the efficiency-accuracy trade-off.
  5. [Sec. 1 and Sec. 4.4] Several typos remain, including 'multigate' for 'mitigate' and 'muti-faceted' for 'multi-faceted'; a careful proofread is needed.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the SALCR pipeline is a standard unsupervised self-training loop, and the central claims are validated on held-out test data against external baselines.

full rationale

The paper's derivation chain is a conventional unsupervised pseudo-label learning pipeline: intra-modality clustering produces pseudo-labels, a dual-association step (OTLA, Eq. 4) unifies label spaces across modalities, and contrastive losses (Eqs. 5-8, 12-15, 24-27) optimize global and part features with respect to those labels. This is self-referential in the same way all self-training methods are, but it is not circular in the sense targeted here: the pseudo-labels are generated from training features, and the reported Rank-1/mAP numbers are evaluated on held-out test sets with ground-truth identity correspondence, not on the training objective itself. The FGSAL 'potential positive pair' criterion (shared pseudo-label in either label space) defines a training target rather than a prediction, and the paper supports the module's contribution with controlled ablations (Table 3) and external comparisons. The GPCR positive mining similarly defines an auxiliary contrastive objective, not a fitted quantity that is later reported as a result. CMFP is a post-processing/re-ranking step applied to features; it introduces hyper-parameters (ktr, kte) tuned on validation behavior, but this is standard hyper-parameter selection, not a fitted input renamed as a prediction. The paper's self-citations (MBCCM, DOTLA, MULT) appear as related-work baselines and comparison methods; none supplies a load-bearing uniqueness theorem, an ansatz smuggled in by citation, or a premise the central result reduces to. No equation in the paper is equivalent by construction to the performance claim, and no cited author-derived result is used to forbid alternatives. The appropriate finding is therefore no significant circularity.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

The framework relies on standard ML components (optimal transport, contrastive learning, clustering) and domain-specific assumptions about the semantic alignment of body parts across modalities. Hyperparameters are tuned per dataset without explicit theoretical justification. No new physical entities or mediators are introduced.

free parameters (6)
  • Np (number of parts) = 3
    Tuned via parameter analysis (Tab. 9-10); chosen as the best value on both datasets, attributed to head/upper/lower body division.
  • lambda (loss weight for GPCR) = 0.5
    Tuned via parameter analysis (Fig. 5-6); chosen as the best trade-off between cluster-level and instance-level losses.
  • k (number of neighbors in GPCR) = 30
    Tuned via parameter analysis (Fig. 5-6); chosen as the best on SYSU-MM01, same value applied to RegDB.
  • ktr (neighbors for CMFP during training) = 30 for SYSU-MM01, 8 for RegDB
    Tuned via parameter analysis (Tab. 11-12); chosen per dataset based on scale.
  • kte (neighbors for CMFP during testing) = 30 for SYSU-MM01, 8 for RegDB
    Tuned via parameter analysis (Tab. 13-14); chosen per dataset.
  • DBSCAN eps = 0.6 for SYSU-MM01, 0.3 for RegDB
    Standard hyperparameter inherited from DCL baseline; mentioned in implementation details, not tuned in this paper.
assumptions (4)
  • standard math Sinkhorn-Knopp algorithm solves the optimal transport problem in Eq. 4.
    Invoked in Sec. 3.2 to compute the transport plan for dual association; no proof given, used as a known tool.
  • domain assumption InfoNCE contrastive loss is effective for learning identity-discriminative features.
    Used throughout (Eq. 1-2, 5-6, 12-13, etc.); standard in ReID literature, but its effective use with noisy pseudo-labels is assumed.
  • domain assumption Dividing the feature map into Np=3 parts corresponds to meaningful semantic body parts.
    Introduced in Sec. 3.3 and used in FGSAL; the paper states this is 'an intuitive and reasonable strategy' but does not verify semantic alignment per part.
  • domain assumption Cross-modality pairs sharing a pseudo-label are likely true positives.
    In Sec. 3.3, potential positive pairs are defined by shared pseudo-labels; this is the basis for FGSAL and GPCR, and its reliability is not quantified.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Semantic-Aligned Learning with Collaborative Refinement for Unsupervised VI-ReID." pith.science (2026). https://pith.science/paper/DOWFKYLB

@misc{pith2026250419244,
  author       = {Pith},
  title        = {Pith review of: Semantic-Aligned Learning with Collaborative Refinement for Unsupervised VI-ReID},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DOWFKYLB}},
  note         = {Machine review of arXiv:2504.19244}
}
read the original abstract

Unsupervised visible-infrared person re-identification (USL-VI-ReID) seeks to match pedestrian images of the same individual across different modalities without human annotations for model learning. Previous methods unify pseudo-labels of cross-modality images through label association algorithms and then design contrastive learning framework for global feature learning. However, these methods overlook the cross-modality variations in feature representation and pseudo-label distributions brought by fine-grained patterns. This insight results in insufficient modality-shared learning when only global features are optimized. To address this issue, we propose a Semantic-Aligned Learning with Collaborative Refinement (SALCR) framework, which builds up optimization objective for specific fine-grained patterns emphasized by each modality, thereby achieving complementary alignment between the label distributions of different modalities. Specifically, we first introduce a Dual Association with Global Learning (DAGI) module to unify the pseudo-labels of cross-modality instances in a bi-directional manner. Afterward, a Fine-Grained Semantic-Aligned Learning (FGSAL) module is carried out to explore part-level semantic-aligned patterns emphasized by each modality from cross-modality instances. Optimization objective is then formulated based on the semantic-aligned features and their corresponding label space. To alleviate the side-effects arising from noisy pseudo-labels, we propose a Global-Part Collaborative Refinement (GPCR) module to mine reliable positive sample sets for the global and part features dynamically and optimize the inter-instance relationships. Extensive experiments demonstrate the effectiveness of the proposed method, which achieves superior performances to state-of-the-art methods. Our code is available at \href{https://github.com/FranklinLingfeng/code-for-SALCR}.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

71 extracted references · 60 canonical work pages

  1. [1]

    In: European Conference on Computer Vision, Springer, pp 720--737

    Alehdaghi M, Josi A, Cruz RM, Granger E (2022) Visible-infrared person re-identification using privileged intermediate information. In: European Conference on Computer Vision, Springer, pp 720--737

    work page 2022

  2. [2]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp 14960--14969

    Chen H, Lagadec B, Bremond F (2021) Ice: Inter-instance contrastive encoding for unsupervised person re-identification. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp 14960--14969

    work page 2021

  3. [3]

    Chen Z, Zhang Z, Tan X, Qu Y, Xie Y (2023) Unveiling the power of clip in unsupervised visible-infrared person re-identification. In: Proceedings of the 31st ACM International Conference on Multimedia, Association for Computing Machinery, New York, NY, USA, MM '23, p 3667–3675, doi:10.1145/3581783.3612050, ://doi.org/10.1145/3581783.3612050

    arXiv 2023

  4. [4]

    In: Proceedings of the 31st ACM International Conference on Multimedia, pp 1325--1333

    Cheng D, He L, Wang N, Zhang S, Wang Z, Gao X (2023 a ) Efficient bilateral cross-modality cluster matching for unsupervised visible-infrared person reid. In: Proceedings of the 31st ACM International Conference on Multimedia, pp 1325--1333

    work page 2023

  5. [5]

    In: Proceedings of the 31st ACM International Conference on Multimedia, pp 7085--7093

    Cheng D, Huang X, Wang N, He L, Li Z, Gao X (2023 b ) Unsupervised visible-infrared person reid by collaborative learning with neighbor-guided label refinement. In: Proceedings of the 31st ACM International Conference on Multimedia, pp 7085--7093

    work page 2023

  6. [6]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 7308--7318

    Cho Y, Kim WJ, Hong S, Yoon SE (2022) Part-based pseudo label refinement for unsupervised person re-identification. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 7308--7318

    work page 2022

  7. [7]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 10257--10266

    Choi S, Lee S, Kim Y, Kim T, Kim C (2020) Hi-cmd: Hierarchical cross-modality disentanglement for visible-infrared person re-identification. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 10257--10266

    work page 2020

  8. [8]

    Advances in neural information processing systems 26

    Cuturi M (2013) Sinkhorn distances: Lightspeed computation of optimal transport. Advances in neural information processing systems 26

    work page 2013

Show all 71 references
  1. [9]

    In: IJCAI, vol 1, p 6

    Dai P, Ji R, Wang H, Wu Q, Huang Y (2018) Cross-modality person re-identification with generative adversarial training. In: IJCAI, vol 1, p 6

    2018

  2. [10]

    In: Proceedings of the Asian Conference on Computer Vision, pp 1142--1160

    Dai Z, Wang G, Yuan W, Zhu S, Tan P (2022) Cluster contrast for unsupervised person re-identification. In: Proceedings of the Asian Conference on Computer Vision, pp 1142--1160

    2022

  3. [11]

    In: 2009 IEEE conference on computer vision and pattern recognition, Ieee, pp 248--255

    Deng J, Dong W, Socher R, Li LJ, Li K, Fei-Fei L (2009) Imagenet: A large-scale hierarchical image database. In: 2009 IEEE conference on computer vision and pattern recognition, Ieee, pp 248--255

    2009

  4. [12]

    In: kdd, vol 96, pp 226--231

    Ester M, Kriegel HP, Sander J, Xu X, et al (1996) A density-based algorithm for discovering clusters in large spatial databases with noise. In: kdd, vol 96, pp 226--231

    1996

  5. [13]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp 11270--11279

    Fang X, Yang Y, Fu Y (2023) Visible-infrared person re-identification via semantic alignment and affinity inference. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp 11270--11279

    2023

  6. [14]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp 22752--22761

    Feng J, Wu A, Zheng WS (2023) Shape-erased feature learning for visible-infrared person re-identification. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp 22752--22761

    2023

  7. [15]

    In: Proceedings of the 1998 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP'98 (Cat

    Frigo M, Johnson SG (1998) Fftw: An adaptive software architecture for the fft. In: Proceedings of the 1998 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP'98 (Cat. No. 98CH36181), IEEE, vol 3, pp 1381--1384

    1998

  8. [16]

    In: International Conference on Learning Representations, ://openreview.net/forum?id=rJlnOhVYPS

    Ge Y, Chen D, Li H (2020 a ) Mutual mean-teaching: Pseudo label refinery for unsupervised domain adaptation on person re-identification. In: International Conference on Learning Representations, ://openreview.net/forum?id=rJlnOhVYPS

    2020

  9. [17]

    Advances in neural information processing systems 33:11309--11321

    Ge Y, Zhu F, Chen D, Zhao R, et al (2020 b ) Self-paced contrastive learning with hybrid memory for domain adaptive object re-id. Advances in neural information processing systems 33:11309--11321

    2020

  10. [18]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 770--778

    He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 770--778

    2016

  11. [19]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 9729--9738

    He K, Fan H, Wu Y, Xie S, Girshick R (2020) Momentum contrast for unsupervised visual representation learning. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 9729--9738

    2020

  12. [20]

    International Journal of Computer Vision pp 1--20

    He L, Cheng D, Wang N, Gao X (2024) Exploring homogeneous and heterogeneous consistent label associations for unsupervised visible-infrared person reid. International Journal of Computer Vision pp 1--20

    2024

  13. [21]

    In: European Conference on Computer Vision, Springer, pp 480--496

    Jiang K, Zhang T, Liu X, Qian B, Zhang Y, Wu F (2022) Cross-modality transformer for visible-infrared person re-identification. In: European Conference on Computer Vision, Springer, pp 480--496

    2022

  14. [22]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp 18621--18632

    Kim M, Kim S, Park J, Park S, Sohn K (2023) Partmix: Regularization strategy to learn part discovery for visible-infrared person re-identification. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp 18621--18632

    2023

  15. [23]

    In: European Conference on Computer Vision, Springer, pp 381--398

    Li X, Lu Y, Liu B, Liu Y, Yin G, Chu Q, Huang J, Zhu F, Zhao R, Yu N (2022) Counterfactual intervention feature transfer for visible-infrared person re-identification. In: European Conference on Computer Vision, Springer, pp 381--398

    2022

  16. [24]

    arXiv preprint arXiv:240101839

    Li Y, Zhang T, Zhang Y (2024) Frequency domain modality-invariant feature learning for visible-infrared person re-identification. arXiv preprint arXiv:240101839

    2024

  17. [25]

    IEEE Transactions on Circuits and Systems for Video Technology

    Liang T, Jin Y, Liu W, Wang T, Feng S, Li Y (2024) Bridging the gap: Multi-level cross-modality joint alignment for visible-infrared person re-identification. IEEE Transactions on Circuits and Systems for Video Technology

    2024

  18. [26]

    IEEE Transactions on Image Processing 30:6392--6407

    Liang W, Wang G, Lai J, Xie X (2021) Homogeneous-to-heterogeneous: Unsupervised learning for rgb-infrared person re-identification. IEEE Transactions on Image Processing 30:6392--6407

    2021

  19. [27]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp 19366--19375

    Liu J, Sun Y, Zhu F, Pei H, Yang Y, Li W (2022) Learning memory-augmented unidirectional metrics for cross-modality person re-identification. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp 19366--19375

    2022

  20. [28]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

    Lu Y, Wu Y, Liu B, Zhang T, Li B, Chu Q, Yu N (2020) Cross-modality person re-identification with shared-specific feature transfer. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

    2020

  21. [29]

    In: Proceedings of the IEEE/CVF international conference on computer vision, pp 4976--4985

    Luo C, Chen Y, Wang N, Zhang Z (2019) Spectral feature transformation for person re-identification. In: Proceedings of the IEEE/CVF international conference on computer vision, pp 4976--4985

    2019

  22. [30]

    Journal of machine learning research 9(11)

    Van der Maaten L, Hinton G (2008) Visualizing data using t-sne. Journal of machine learning research 9(11)

    2008

  23. [31]

    1707.01400

    Mao X, Li Q, Xie H (2017) Aligngan: Learning to align cross-domain images with conditional generative adversarial networks. 1707.01400

    2017 arXiv

  24. [32]

    Sensors 17(3):605

    Nguyen DT, Hong HG, Kim KW, Park KR (2017) Person recognition system based on a combination of body images from visible light and thermal cameras. Sensors 17(3):605

    2017

  25. [33]

    IEEE Transactions on Circuits and Systems for Video Technology pp 1--1, doi:10.1109/TCSVT.2023.3310015

    Pang Z, Wang C, Zhao L, Liu Y, Sharma G (2023) Cross-modality hierarchical clustering and refinement for unsupervised visible-infrared person re-identification. IEEE Transactions on Circuits and Systems for Video Technology pp 1--1, doi:10.1109/TCSVT.2023.3310015

    2023

  26. [34]

    Knowledge-Based Systems 285:111350, doi:https://doi.org/10.1016/j.knosys.2023.111350, ://www.sciencedirect.com/science/article/pii/S0950705123010985

    Pang Z, Wang C, Pan H, Zhao L, Wang J, Guo M (2024 a ) Mimr: Modality-invariance modeling and refinement for unsupervised visible-infrared person re-identification. Knowledge-Based Systems 285:111350, doi:https://doi.org/10.1016/j.knosys.2023.111350, ://www.sciencedirect.com/s...

    2024

  27. [35]

    IEEE Transactions on Circuits and Systems for Video Technology

    Pang Z, Zhao L, Liu Y, Sharma G, Wang C (2024 b ) Inter-modality similarity learning for unsupervised multi-modality person re-identification. IEEE Transactions on Circuits and Systems for Video Technology

    2024

  28. [36]

    In: International conference on machine learning, PMLR, pp 8748--8763

    Radford A, Kim JW, Hallacy C, Ramesh A, Goh G, Agarwal S, Sastry G, Askell A, Mishkin P, Clark J, et al (2021) Learning transferable visual models from natural language supervision. In: International conference on machine learning, PMLR, pp 8748--8763

    2021

  29. [37]

    arXiv preprint arXiv:240311708

    Ren K, Zhang L (2024) Implicit discriminative knowledge learning for visible-infrared person re-identification. arXiv preprint arXiv:240311708

    2024

  30. [38]

    In: Proceedings of the IEEE international conference on computer vision, pp 618--626

    Selvaraju RR, Cogswell M, Das A, Vedantam R, Parikh D, Batra D (2017) Grad-cam: Visual explanations from deep networks via gradient-based localization. In: Proceedings of the IEEE international conference on computer vision, pp 618--626

    2017

  31. [39]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp 11218--11228

    Shi J, Zhang Y, Yin X, Xie Y, Zhang Z, Fan J, Shi Z, Qu Y (2023) Dual pseudo-labels interactive self-training for semi-supervised visible-infrared person re-identification. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp 11218--11228

    2023

  32. [40]

    2401.06825

    Shi J, Yin X, Chen Y, Zhang Y, Zhang Z, Xie Y, Qu Y (2024 a ) Multi-memory matching for unsupervised visible-infrared person re-identification. 2401.06825

    2024 arXiv

  33. [41]

    arXiv preprint arXiv:240219026

    Shi J, Yin X, Wang Y, Liu X, Xie Y, Qu Y (2024 b ) Progressive contrastive learning with multi-prototype for unsupervised visible-infrared person re-identification. arXiv preprint arXiv:240219026

    2024

  34. [42]

    Information Processing & Management 60(3):103304

    Si T, He F, Li P, Song Y, Fan L (2023) Diversity feature constraint based on heterogeneous data for unsupervised person re-identification. Information Processing & Management 60(3):103304

    2023

  35. [43]

    In: Proceedings of the European conference on computer vision (ECCV), pp 480--496

    Sun Y, Zheng L, Yang Y, Tian Q, Wang S (2018) Beyond part models: Person retrieval with refined part pooling (and a strong convolutional baseline). In: Proceedings of the European conference on computer vision (ECCV), pp 480--496

    2018

  36. [44]

    2302.00884

    Tan L, Zhang Y, Shen S, Wang Y, Dai P, Lin X, Wu Y, Ji R (2023) Exploring invariant representation for visible-infrared person re-identification. 2302.00884

    2023 arXiv

  37. [45]

    In: Proceedings of the 32nd ACM International Conference on Multimedia, pp 9856--9865

    Teng X, Shen X, Xu K, Lan L (2024) Enhancing unsupervised visible-infrared person re-identification with bidirectional-consistency gradual matching. In: Proceedings of the 32nd ACM International Conference on Multimedia, pp 9856--9865

    2024

  38. [46]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp 1522--1531

    Tian X, Zhang Z, Lin S, Qu Y, Xie Y, Ma L (2021) Farewell to mutual information: Variational distillation for cross-modal person re-identification. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp 1522--1531

    2021

  39. [47]

    Advances in neural information processing systems 30

    Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez AN, Kaiser , Polosukhin I (2017) Attention is all you need. Advances in neural information processing systems 30

    2017

  40. [48]

    In: European Conference on Computer Vision, Springer, pp 93--109

    Wang J, Zhang Z, Chen M, Zhang Y, Wang C, Sheng B, Qu Y, Xie Y (2022 a ) Optimal transport for label-efficient visible-infrared person re-identification. In: European Conference on Computer Vision, Springer, pp 93--109

    2022

  41. [49]

    IEEE Transactions on Image Processing 31:6548--6561

    Wang M, Li J, Lai B, Gong X, Hua XS (2022 b ) Offline-online associated camera-aware proxies for unsupervised person re-identification. IEEE Transactions on Image Processing 31:6548--6561

    2022

  42. [50]

    In: Proceedings of the IEEE international conference on computer vision, pp 5380--5389

    Wu A, Zheng WS, Yu HX, Gong S, Lai J (2017) Rgb-infrared cross-modality person re-identification. In: Proceedings of the IEEE international conference on computer vision, pp 5380--5389

    2017

  43. [51]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp 11122--11131

    Wu J, Liu H, Su Y, Shi W, Tang H (2023) Learning concordant attention via target-aware alignment for visible-infrared person re-identification. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp 11122--11131

    2023

  44. [52]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp 4330--4339

    Wu Q, Dai P, Chen J, Lin CW, Wu Y, Huang F, Zhong B, Ji R (2021) Discover cross-modality nuances for visible-infrared person re-identification. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp 4330--4339

    2021

  45. [53]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp 9548--9558

    Wu Z, Ye M (2023) Unsupervised visible-infrared person re-identification via progressive graph matching and alternate learning. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp 9548--9558

    2023

  46. [54]

    Yang B, Ye M, Chen J, Wu Z (2022) Augmented dual-contrastive aggregation learning for unsupervised visible-infrared person re-identification. In: Proceedings of the 30th ACM International Conference on Multimedia, Association for Computing Machinery, New York, NY, USA, MM '22,...

    2022

  47. [55]

    IEEE Transactions on Information Forensics and Security

    Yang B, Chen J, Chen C, Ye M (2023 a ) Dual consistency-constrained learning for unsupervised visible-infrared person re-identification. IEEE Transactions on Information Forensics and Security

    2023

  48. [56]

    IEEE Transactions on Image Processing

    Yang B, Chen J, Ma X, Ye M (2023 b ) Translation, association and augmentation: Learning cross-modality re-identification from single-modality annotation. IEEE Transactions on Image Processing

    2023

  49. [57]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pp 11069--11079

    Yang B, Chen J, Ye M (2023 c ) Towards grand unified representation learning for unsupervised visible-infrared person re-identification. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pp 11069--11079

    2023

  50. [58]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp 16870--16879

    Yang B, Chen J, Ye M (2024) Shallow-deep collaborative learning for unsupervised visible-infrared person re-identification. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp 16870--16879

    2024

  51. [59]

    IEEE Transactions on Information Forensics and Security

    Yang Y, Hu W, Hu H (2025) Progressive cross-modal association learning for unsupervised visible-infrared person re-identification. IEEE Transactions on Information Forensics and Security

    2025

  52. [60]

    In: Computer Vision--ECCV 2020: 16th European Conference, Glasgow, UK, August 23--28, 2020, Proceedings, Part XVII 16, Springer, pp 229--247

    Ye M, Shen J, J Crandall D, Shao L, Luo J (2020) Dynamic dual-attentive aggregation learning for visible-infrared person re-identification. In: Computer Vision--ECCV 2020: 16th European Conference, Glasgow, UK, August 23--28, 2020, Proceedings, Part XVII 16, Springer, pp 229--247

    2020

  53. [61]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp 13567--13576

    Ye M, Ruan W, Du B, Shou MZ (2021 a ) Channel augmented joint learning for visible-infrared recognition. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp 13567--13576

    2021

  54. [62]

    2001.04193

    Ye M, Shen J, Lin G, Xiang T, Shao L, Hoi SCH (2021 b ) Deep learning for person re-identification: A survey and outlook. 2001.04193

    2021 arXiv

  55. [63]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp 7349--7358

    Zhang Q, Lai C, Liu J, Huang N, Han J (2022 a ) Fmcnet: Feature-level modality compensation for visible-infrared person re-identification. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp 7349--7358

    2022

  56. [64]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp 3436--3445

    Zhang X, Ge Y, Qiao Y, Li H (2021) Refining pseudo labels with clustering consensus over generations for unsupervised object re-identification. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp 3436--3445

    2021

  57. [65]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp 7369--7378

    Zhang X, Li D, Wang Z, Wang J, Ding E, Shi JQ, Zhang Z, Wang J (2022 b ) Implicit sample extension for unsupervised person re-identification. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp 7369--7378

    2022

  58. [66]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp 2153--2162

    Zhang Y, Wang H (2023) Diverse embedding expansion network and low-light cross-modality benchmark for visible-infrared person re-identification. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp 2153--2162

    2023

  59. [67]

    arXiv preprint arXiv:240102162

    Zhang Y, Lu Y, Yan Y, Wang H, Li X (2024) Frequency domain nuances mining for visible-infrared person re-identification. arXiv preprint arXiv:240102162

    2024

  60. [68]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 1318--1327

    Zhong Z, Zheng L, Cao D, Li S (2017) Re-ranking person re-identification with k-reciprocal encoding. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 1318--1327

    2017

  61. [69]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pp 11058--11068

    Zou C, Chen Z, Cui Z, Liu Y, Zhang C (2023) Discrepant and multi-instance proxies for unsupervised person re-identification. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pp 11058--11068

    2023

  62. [70]

    , " * write output.state after.block = add.period write newline

    ENTRY address archive author booktitle chapter doi edition editor eid eprint howpublished institution journal key month note number organization pages publisher school series title type url volume year label extra.label sort.label short.list INTEGERS output.state before.all mi...

  63. [71]

    write newline

    " write newline "" before.all 'output.state := FUNCTION add.period duplicate empty 'skip "." * add.blank if FUNCTION if.digit duplicate "0" = swap duplicate "1" = swap duplicate "2" = swap duplicate "3" = swap duplicate "4" = swap duplicate "5" = swap duplicate "6" = swap dupl...

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.