Reliability-Aware 3D Geometric Injection for Universal Person Re-identification

Bohan Su; Fangyi Liu; Jiashuo Wang; Mang Ye

REVIEW 3 major objections 5 minor 106 references

Reliability-Aware 3D Geometric Injection for Universal Person Re-identification

T0 review · 3 major / 5 minor · reviewed 2026-08-01 · deepseek-v4-flash

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

Pith's one-line read A learned reliability gate makes monocular 3D pose a safe, conditional cue for universal person re-identification, improving challenging scenarios while preserving clean-domain performance.

desk verdict Useful engineering recipe for 3D-injected universal ReID, but the learned gate's claim to measure per-sample reliability is not backed by the evidence. read the letter →

arxiv 2607.18863 v1 pith:QY4ONVQQ submitted 2026-07-21 cs.CV

Bohan Su , Jiashuo Wang , Fangyi Liu , Mang Ye This is my paper

classification cs.CV

keywords universalpersonre-identificationmonocular3DpriorsreliabilitygateresidualfusionSMPLnegativetransferocclusioncross-modality

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

The paper claims that monocular 3D body geometry can help a single person re-identification model across diverse scenarios—occlusion, clothing change, cross-modality—if it is injected conditionally rather than unconditionally. Naively fusing 3D estimates often hurts because the estimates are noisy under extreme degradation. The proposed solution predicts a per-sample reliability scalar from the consistency between 2D appearance and 3D pose, and adds the 3D feature only as a residual scaled by that scalar. This lets the model fall back to pure 2D when geometry is unreliable, which the authors show yields gains of +3.0% Rank-1 on PRCC, +0.9% on Occluded-Duke, and +1.2% on SYSU-MM01 while staying competitive on clean benchmarks. If true, it turns 3D geometry into a dependable auxiliary channel for retrieval.

What carries the argument

The load-bearing object is the Consistency-Aware Reliability Gate: a bottleneck MLP that maps the concatenated visual and pose features to a single sigmoid scalar α. This α modulates a dual-stream residual fusion, f_out = [f_vis; f_vis + α·f_pose], isolating the 3D prior as a late-stage structural residual. The pose feature itself is produced by a kinematic-aware encoder that decouples SMPL parameters into global and local joint tokens, processes them with a small Transformer, and pools only the refined joint tokens to form a view-invariant geometric representation.

What would settle it

Take one clean dataset (e.g., Market-1501), run the trained model on original images and on the same images with synthetic heavy occlusion or blur, and compare the predicted α. If α does not drop for the corrupted images and retrieval gains do not track per-sample variations, the gate is not filtering geometric noise at the sample level.

Watch

Extended reading notes

Core claim

On its own terms, the paper establishes that 3D geometry should be treated as conditional structural evidence, not a uniformly trusted input. The core mechanism is a late residual fusion: the final descriptor is [f_vis; f_vis + α·f_pose], where f_vis is the 2D visual feature, f_pose is a kinematic joint representation extracted from SMPL parameters, and α = σ(MLP([f_vis; f_pose])) is a learned scalar in (0,1). When α≈0, the representation reduces to the pure 2D baseline, providing a controlled fallback. The paper demonstrates that this gated fusion improves structure-sensitive scenarios while avoiding the negative transfer that occurs with unconditional fusion.

Load-bearing premise

The paper's claim depends on the learned scalar α measuring genuine per-sample geometric reliability; if α instead learns dataset-level difficulty from the scene prompts, the observed gains could be reproduced by a single constant scaling factor tuned per benchmark.

Editorial extensions

If this is right

Universal ReID models can incorporate monocular 3D geometry without per-dataset retuning, since the gate learns when to suppress unreliable estimates.
The late-residual design provides a built-in deployment fallback: when 3D extraction is unavailable, forcing α=0 yields a descriptor whose cosine similarity is equivalent to the pure 2D baseline.
The gate's behavior suggests that auxiliary modalities in general should be fused proportionally to estimated reliability rather than as equally trusted signals.
The kinematic decoupling of SMPL parameters indicates that representing body structure as joint tokens, not a flat vector, is important for stable optimization.

Reading between the lines

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

The reported α distribution (≈0.15 on clean datasets vs ≈0.82 on clothing-change) raises the possibility that α mostly encodes a dataset-level switch rather than per-sample geometric quality; testing on corrupted samples within one clean dataset would distinguish these.
Because the fusion is a feature-space residual, the gating idea could transfer to other retrieval tasks where an auxiliary channel (depth, skeleton, text) is reliable only part of the time.
The authors leave open explicit supervision of α with geometric quality scores (e.g., keypoint confidence); adding such a signal could sharpen per-sample gating and reduce calibration dependence across benchmarks.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Short take: the paper gets a real result — late residual fusion with a learned scalar gate beats both the pure 2D baseline and naive 3D concat on the benchmarks that matter, and it does so without wrecking clean-domain performance. But the paper's central interpretation, that the gate is detecting per-sample geometric reliability, is not established. The evidence is equally consistent with the gate learning a dataset-level switch.

What's new: the kinematic-aware decoupling of SMPL parameters (global vs joint rotation, cross-attention with global, pooling only joints) is a sensible, clearly explained design, and the ablation in Table 5 shows it matters. The dual-stream residual fusion with fallback to pure 2D is a clean and useful deployment trick. The experimental coverage is genuinely extensive: 9 datasets, five scenario groups, transfer protocols, and a comparison against three stronger fusion alternatives. The paper is honest about its limits — the conclusion explicitly says correlating the gate with explicit geometric quality metrics is future work.

Where it's soft: the load-bearing claim is that the scalar alpha filters per-sample geometric noise. Alpha is trained with only identity losses; there is no auxiliary signal tying it to 3D quality. The only analysis of alpha, Fig. 3, shows strong clustering by scenario group (mean 0.15 in Group A vs 0.82 in Group B), which is exactly what you'd expect if alpha were a per-dataset or per-prompt bias rather than a per-sample reliability detector. No within-dataset correlation between alpha and any geometric error is reported. So the 'consistency-aware reliability' framing overstates what is actually learned. The gains are also modest (0.4-1.2% R1 except PRCC +3.0%), and all numbers are single-run point estimates with no error bars. Table 5 compares the black-box SMPL variant at 120 epochs with everything else at 180, and Table 6 reports HyperNet Prompt at 60 epochs; that's a fairness smell even if the authors explain why.

That said, the central engineering result — gated late residual is safe and helps — survives the criticism. The negative-transfer story holds empirically: the model does not degrade clean benchmarks. The gap is interpretation, not existence.

Who this is for: anyone working on universal ReID or multi-modal fusion with noisy auxiliary inputs. It deserves a serious referee. I'd recommend the editor send it out, with the understanding that the reliability-gating claim needs stronger per-sample validation before the paper can be accepted as-is.

Referee Report

3 major / 5 minor

Summary. The paper proposes UniGeo, a universal person ReID framework that augments a ViT-based 2D visual stream with monocular SMPL geometry via a kinematic-aware pose encoder. A reliability gate α predicted by an MLP from concatenated visual and geometric features modulates a late residual fusion, f_out = [f_vis; f_vis + α·f_pose]. The model is trained jointly on multiple datasets with standard classification and triplet losses, and evaluated on nine benchmarks spanning holistic, clothing-change, occlusion, cross-modality, and UAV/wild settings. The central claim is that the learned gate detects per-sample visual-geometric inconsistency and suppresses unreliable 3D cues, yielding gains on structure-sensitive benchmarks without negative transfer on clean ones.

Significance. The paper addresses a real problem and the architecture is clean: decoupling 3D extraction from utilization, preserving a pure-2D fallback, and using a scalar gate with only ~0.3M extra parameters is attractive. Strengths include a controlled comparison against a shared pure-2D baseline, multiple ablations (Tables 4–6), and release of code. If the gate truly filters per-sample geometric noise, this would be a useful contribution to universal ReID. However, the central mechanism is not yet empirically established; the current evidence is consistent with dataset-level bias.

major comments (3)

[Sec. 3.3, Eqs. (4) and (6); Sec. 4.4] The reliability interpretation of α is load-bearing, but α is trained only by identity classification and triplet losses; no auxiliary signal ties α to geometric estimation quality. Sec. 4.4 reports only group-level means (Group A 0.15±0.05 vs Group B 0.82±0.10). Since f_vis includes dataset-conditioned scene prompts (Sec. 3.1), these group differences could reflect dataset-level bias rather than per-sample consistency. The paper explicitly defers correlation with explicit geometric quality metrics to future work. Please provide within-dataset per-sample analyses (e.g., correlation of α with SMPL fitting loss or 2D keypoint reprojection error) or an oracle-gating comparison; without this, the claimed reliability-aware mechanism is not distinguishable from a per-dataset scaling factor.
[Tables 1–6] All reported numbers are single-run point estimates. The claimed gains are small (e.g., +0.4 avg mAP in Table 6, +0.9 Rank-1 on Occluded-Duke in Table 2b), so the universal-improvement claim is not robust without variance estimates. Please report mean±std over at least three seeds and run paired significance tests for the key baseline-vs-UniGeo comparisons.
[Sec. 4.4, Table 2(b), Sec. 4.1] SYSU-MM01 is a cross-modality benchmark, but the SMPL extractor 4DHumans is RGB-trained. The paper reports a high mean α of 0.68 for this group, implying strong geometric injection, yet provides no evidence about the quality of SMPL estimates on infrared images. This is a concrete instance of the gate-validation gap: either report geometric quality metrics on infrared inputs or explain how high α is compatible with the noise-suppression narrative.

minor comments (5)

[Table 3] The row/column alignment for the four AG-ReID.v2 transfer protocols is hard to read; use explicit column separators and clarify which metric corresponds to each protocol.
[Eq. (3)] Notation is inconsistent: the text defines T_J but uses t_{J,j}; please align the notation for joint tokens.
[Sec. 3.4] The statement that [f_vis; f_vis] is retrieval-equivalent to f_vis holds for cosine similarity at inference, but training uses the duplicated vector in a classifier; clarify that the equivalence is an inference-time property.
[Fig. 3] If the figure shows only group means, replace or augment it with per-sample α distributions or scatter plots; group means alone are insufficient to support the per-sample reliability claim.
[Table 6, footnote] The HyperNet Prompt baseline is evaluated at epoch 60 due to training instability; state this limitation in the main text rather than only in a footnote.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: test-set gains are measured, not derived; the unvalidated reliability interpretation is a validity caveat, not a circular step.

full rationale

The paper's derivation chain is empirical rather than analytic. The final representation f_out in Eq. (5) is a design choice; its test-set gains (Tables 2, 4) are measured against held-out data. No equation in the paper defines a benchmark result as equal to a fitted parameter. The gate α is trained with the same identity/triplet losses as the rest of the model, but the paper does not claim to derive α from an independent reliability signal; it claims the gate behaves as a reliability filter. That interpretation is not independently validated — the paper itself says "Future work will correlate the learned gate with explicit geometric quality metrics" — which is a limitation of construct validity, not circularity: if the gate were only a dataset-level switch, the reported gains could still be genuine, only differently attributed. The controlled comparison against a pure-2D baseline and the Naive-3D ablation show the gated result is not force-fitted: naive concat regresses on PRCC, so the +3.0% is not an artifact of construction. Self-citations (e.g., Refs. [47,48]) appear only in related work and are not load-bearing. No pattern of circularity is present.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The method rests on engineering assumptions rather than mathematical axioms. The key unproven premises are that SMPL estimates are informative under extreme degradation, that a learned scalar gate can measure reliability without explicit supervision, and that late residual fusion is safe. These are plausible but not independently established.

free parameters (5)

Scene-aware prompt set (per-dataset prompts, length K) = K unspecified; prompt tokens learned per dataset
Datasets are assigned semantic scene labels; prompt conditioning is a learnable per-dataset parameter that already gives the model dataset identity, making the later gate's dataset-level activation pattern non-surprising.
Pose Transformer hyperparameters = 2 layers, hidden 768, 8 heads
Chosen by hand; no sensitivity analysis reported.
Gate MLP architecture = 1536 -> 768 -> 1
Chosen by hand; no sensitivity analysis reported.
Training hyperparameters = 180 epochs, SGD lr=4e-4, batch 120, triplet margin 0.3, lambda=1
Standard ReID recipe; not fitted to target data but not justified by analysis.
Fusion design choice: global params used via cross-attention but discarded from pooling = Full topology (s_global + s_joint), joint-token pooling
Selected via Table 5 ablation on test sets; a free design choice, not derived from first principles.

assumptions (4)

domain assumption SMPL parameters estimated by frozen 4DHumans from monocular RGB images remain informative under occlusion, clothing change, and RGB-to-IR shift.
The entire 3D branch rests on this; the paper cites 4DHumans but provides no validation on these degraded domains.
ad hoc to paper The scalar gate α, trained only by ReID losses, is a valid proxy for geometric reliability.
Eq. (4) defines α as an MLP over concatenated f_vis and f_pose; no external reliability labels or geometric quality losses are used. Fig. 3 shows dataset-level variation, not per-sample validation.
domain assumption Late residual fusion with α∈[0,1] cannot harm the 2D baseline and preserves retrieval behavior when α=0.
Eq. (5) and Sec. 3.4 argue that α=0 yields [f_vis; f_vis] with cosine similarity equal to pure 2D; true for the retrieval metric, but training dynamics with α>0 are not analyzed.
domain assumption The chosen public benchmark splits and joint-training protocol are appropriate for measuring universal ReID.
Standard practice, but the universal baseline is an extension of VersReID and is not independently validated.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Reliability-Aware 3D Geometric Injection for Universal Person Re-identification." pith.science (2026). https://pith.science/paper/QY4ONVQQ

@misc{pith2026260718863,
  author       = {Pith},
  title        = {Pith review of: Reliability-Aware 3D Geometric Injection for Universal Person Re-identification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QY4ONVQQ}},
  note         = {Machine review of arXiv:2607.18863}
}

read the original abstract

Universal person re-identification (ReID) aims to retrieve pedestrian identities across diverse real-world scenarios, including severe occlusions, clothing changes, and cross-modality shifts, within a unified model. However, existing 2D representations fundamentally struggle with spatial ambiguities due to a lack of depth and topological awareness, while naively introducing monocular 3D priors often causes severe negative transfer due to geometric estimation noise under extreme visual degradation. To safely harness the clothing-invariant and canonical structural properties of 3D geometry, we propose UniGeo, a Universal Monocular 3D-Enhanced ReID framework driven by a Consistency-Aware Reliability Gate and Dual-Stream Residual Fusion. Specifically, the processing of 3D information is strategically decoupled into geometric extraction and dynamic utilization. To provide pure structural compensation, we project monocular 3D parameters into kinematic joint representations, explicitly capturing instance-level geometric topology to resolve appearance-based ambiguities. To robustly incorporate these cues without perturbing the reliable 2D feature space, we isolate the 3D prior as a late-stage structural residual; modulated by the consistency-aware gate, this mechanism adaptively filters geometric noise and enables controlled fallback to the pure 2D baseline. Extensive experiments show that our method improves challenging, structure-sensitive scenarios while preserving competitive performance on clean domains. Code is available at https://github.com/BohanSu/UniGeo.

Figures

Figures reproduced from arXiv: 2607.18863 by the authors.

**Figure 1.** (a) Standard 2D pipelines are vulnerable to severe appearance corruptions (e.g. heavy occlusion, clothing changes). (b) Our 3D-aided pipeline leverages kinematicaware geometry and reliability-aware gating to conditionally fuse 3D structural cues with 2D features, ensuring robust identity representation. dynamically estimates geometric reliability by evaluating cross-modal consistency [60]. This gating mechanism pr… view at source ↗

**Figure 2.** Overview of our UniGeo framework. A Scene-Aware Visual Stream (Sec. 3.1) extracts 2D textures fvis, while an Auxiliary Structural Stream (Sec. 3.2) models decoupled 3D geometric features fpose. Guided by cross-modal discrepancy, a ReliabilityAware Gate predicts scalar α to modulate the dual-stream residual fusion (Sec. 3.3) under end-to-end optimization (Sec. 3.4). assigned a scene label denoting its primary ReID … view at source ↗

**Figure 4.** Training Convergence Analysis showing average mAP trends over epochs during universal joint training. reaching only 72.1% avg mAP at epoch 60. In contrast, Reliability-Gated is the only 3D-augmented variant that improves over the baseline (+0.4% avg mAP, +0.4% avg Rank-1) with only ∼0.3M extra parameters. These stronger alternatives inject geometry too early or too aggressively, forcing noisy structural cues to int… view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

106 extracted references · 5 linked inside Pith

[1]

In: European conference on computer vision

Bogo, F., Kanazawa, A., Lassner, C., Gehler, P., Romero, J., Black, M.J.: Keep it smpl: Automatic estimation of 3d human pose and shape from a single image. In: European conference on computer vision. pp. 561–578. Springer (2016)

2016
[2]

Knowledge-Based Systems307, 112737 (2025)

Che, Q.H., Nguyen, L.C., Luu, D.T., Nguyen, V.T.: Enhancing person re- identification via uncertainty feature fusion method and auto-weighted measure combination. Knowledge-Based Systems307, 112737 (2025)

2025
[3]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Chen, C., Ye, M., Jiang, D.: Towards modality-agnostic person re-identification with descriptive query. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 15128–15137 (2023)

2023
[4]

In: Proceed- ings of the IEEE/CVF conference on computer vision and pattern recognition

Chen, J., Jiang, X., Wang, F., Zhang, J., Zheng, F., Sun, X., Zheng, W.S.: Learn- ing 3d shape feature for texture-insensitive person re-identification. In: Proceed- ings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 8146–8155 (2021)

2021
[5]

In: Proceedings of the IEEE/CVF international conference on computer vision

Chen, T., Ding, S., Xie, J., Yuan, Y., Chen, W., Yang, Y., Ren, Z., Wang, Z.: Abd-net: Attentive but diverse person re-identification. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 8351–8361 (2019)

2019
[6]

In: Proceedings of the IEEE/CVF conference on com- puter vision and pattern recognition

Chen, W., Xu, X., Jia, J., Luo, H., Wang, Y., Wang, F., Jin, R., Sun, X.: Be- yond appearance: a semantic controllable self-supervised learning framework for human-centric visual tasks. In: Proceedings of the IEEE/CVF conference on com- puter vision and pattern recognition. pp. 15050–15061 (2023)

2023
[7]

In: 2024 International Joint Conference on Neural Net- works (IJCNN)

Chen, Z., Ge, Y.: Part-attention based model make occluded person re- identification stronger. In: 2024 International Joint Conference on Neural Net- works (IJCNN). pp. 1–8. IEEE (2024)

2024
[8]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Cheng, Z., Dong, Q., Gong, S., Zhu, X.: Inter-task association critic for cross- resolution person re-identification. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 2605–2615 (2020)

2020

Show all 106 references

[9]

In: European Conference on Computer Vision

Cho, J., Youwang, K., Oh, T.H.: Cross-attention of disentangled modalities for 3d human mesh recovery with transformers. In: European Conference on Computer Vision. pp. 342–359. Springer (2022)

2022
[10]

Choi, S., Lee, S., Kim, Y., Kim, T., Kim, C.: Hi-cmd: Hierarchical cross-modality disentanglementforvisible-infraredpersonre-identification.In:Proceedingsofthe IEEE/CVF conference on computer vision and pattern recognition. pp. 10257– 10266 (2020)

2020
[11]

In: Proceedings of the 32nd ACM International Conference on Multimedia

Cui, C., Huang, S., Song, W., Ding, P., Zhang, M., Wang, D.: Profd: Prompt- guided feature disentangling for occluded person re-identification. In: Proceedings of the 32nd ACM International Conference on Multimedia. pp. 1583–1592 (2024)

2024
[12]

IEEE Transactions on Pattern Analy- sis and Machine Intelligence44(3), 1474–1488 (2020)

Ding, C., Wang, K., Wang, P., Tao, D.: Multi-task learning with coarse priors for robust part-aware person re-identification. IEEE Transactions on Pattern Analy- sis and Machine Intelligence44(3), 1474–1488 (2020)

2020
[13]

IEEE Transactions on Circuits and Systems for Video Technology34(6), 4458–4472 (2023)

Dong, N., Zhang, L., Yan, S., Tang, H., Tang, J.: Erasing, transforming, and noising defense network for occluded person re-identification. IEEE Transactions on Circuits and Systems for Video Technology34(6), 4458–4472 (2023)

2023
[14]

arXiv preprint arXiv:2010.11929 (2020)

Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., et al.: An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929 (2020)

2010 arXiv
[15]

In: Proceedings of the IEEE/CVF conference on com- puter vision and pattern recognition

Feng, J., Wu, A., Zheng, W.S.: Shape-erased feature learning for visible-infrared person re-identification. In: Proceedings of the IEEE/CVF conference on com- puter vision and pattern recognition. pp. 22752–22761 (2023) Universal Monocular 3D-Enhanced ReID 17

2023
[16]

In: Proceedings of the IEEE/CVF international conference on computer vision

Fu, C., Hu, Y., Wu, X., Shi, H., Mei, T., He, R.: Cm-nas: Cross-modality neural architecture search for visible-infrared person re-identification. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 11823–11832 (2021)

2021
[17]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Gao, S., Wang, J., Lu, H., Liu, Z.: Pose-guided visible part matching for occluded person reid. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 11744–11752 (2020)

2020
[18]

In: International conference on learning representations (2018)

Geirhos,R.,Rubisch,P.,Michaelis,C.,Bethge,M.,Wichmann,F.A.,Brendel,W.: Imagenet-trained cnns are biased towards texture; increasing shape bias improves accuracy and robustness. In: International conference on learning representations (2018)

2018
[19]

In: Proceedings of the IEEE/CVF International Conference on Computer Vision

Goel, S., Pavlakos, G., Rajasegaran, J., Kanazawa, A., Malik, J.: Humans in 4d: Reconstructing and tracking humans with transformers. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 14783–14794 (2023)

2023
[20]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Gu, X., Chang, H., Ma, B., Bai, S., Shan, S., Chen, X.: Clothes-changing per- son re-identification with rgb modality only. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 1060–1069 (2022)

2022
[21]

In: Proceedings of the IEEE/CVF international conference on computer vision

He, S., Luo, H., Wang, P., Wang, F., Li, H., Jiang, W.: Transreid: Transformer- based object re-identification. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 15013–15022 (2021)

2021
[22]

In: 2021 IEEE International confer- ence on image processing (ICIP)

Herzog, F., Ji, X., Teepe, T., Hörmann, S., Gilg, J., Rigoll, G.: Lightweight multi- branch network for person re-identification. In: 2021 IEEE International confer- ence on image processing (ICIP). pp. 1129–1133. IEEE (2021)

2021
[23]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Hong, P., Wu, T., Wu, A., Han, X., Zheng, W.S.: Fine-grained shape-appearance mutual learning for cloth-changing person re-identification. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 10513– 10522 (2021)

2021
[24]

In: 2019 International Joint Con- ference on Neural Networks (IJCNN)

Huang, Y., Wu, Q., Xu, J., Zhong, Y.: Celebrities-reid: A benchmark for clothes variation in long-term person re-identification. In: 2019 International Joint Con- ference on Neural Networks (IJCNN). pp. 1–8. IEEE (2019)

2019
[25]

In: Proceedings of the IEEE/CVF interna- tional conference on computer vision

Huang, Y., Wu, Q., Xu, J., Zhong, Y., Zhang, Z.: Clothing status awareness for long-term person re-identification. In: Proceedings of the IEEE/CVF interna- tional conference on computer vision. pp. 11895–11904 (2021)

2021
[26]

IEEE Transactions on Circuits and Systems for Video Technology30(10), 3459–3471 (2019)

Huang, Y., Xu, J., Wu, Q., Zhong, Y., Zhang, P., Zhang, Z.: Beyond scalar neu- ron: Adopting vector-neuron capsules for long-term person re-identification. IEEE Transactions on Circuits and Systems for Video Technology30(10), 3459–3471 (2019)

2019
[27]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Jiang, W., Kolotouros, N., Pavlakos, G., Zhou, X., Daniilidis, K.: Coherent re- construction of multiple humans from a single image. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 5579– 5588 (2020)

2020
[28]

Sensors25(19), 6210 (2025)

Jiang, X., Yuan, X., Yang, X.: Multi-aligned and multi-scale augmentation for occluded person re-identification. Sensors25(19), 6210 (2025)

2025
[29]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Jin, X., He, T., Zheng, K., Yin, Z., Shen, X., Huang, Z., Feng, R., Huang, J., Chen, Z., Hua, X.S.: Cloth-changing person re-identification from a single image with gait prediction and regularization. In: Proceedings of the IEEE/CVF conference on computer vision and pattern re...

2022
[30]

arXiv preprint arXiv:2503.10759 (2025) 18 B

Joseph, A., Peleg, S.: Clothes-changing person re-identification based on skeleton dynamics. arXiv preprint arXiv:2503.10759 (2025) 18 B. Su et al

2025 arXiv
[31]

In: Proceedings of the IEEE conference on computer vision and pattern recognition

Kanazawa, A., Black, M.J., Jacobs, D.W., Malik, J.: End-to-end recovery of hu- man shape and pose. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 7122–7131 (2018)

2018
[32]

Kendall, A., Gal, Y.: What uncertainties do we need in bayesian deep learning for computer vision? Advances in neural information processing systems30(2017)

2017
[33]

Kocabas, M., Huang, C.H.P., Hilliges, O., Black, M.J.: Pare: Part attention regres- sorfor3dhumanbodyestimation.In:ProceedingsoftheIEEE/CVFinternational conference on computer vision. pp. 11127–11137 (2021)

2021
[34]

In: Proceedings of the IEEE/CVF international conference on computer vision

Kolotouros, N., Pavlakos, G., Black, M.J., Daniilidis, K.: Learning to reconstruct 3d human pose and shape via model-fitting in the loop. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 2252–2261 (2019)

2019
[35]

In: Proceedings of the IEEE/CVF international conference on computer vision

Kolotouros, N., Pavlakos, G., Jayaraman, D., Daniilidis, K.: Probabilistic mod- eling for human mesh recovery. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 11605–11614 (2021)

2021
[36]

arXiv preprint arXiv:2510.22268 (2025)

Li, Q., Li, J., Zhang, Y., Tan, L., Chen, J., Ji, J.: Gsalign: Geometric and seman- tic alignment network for aerial-ground person re-identification. arXiv preprint arXiv:2510.22268 (2025)

2025
[37]

Li, S., Sun, L., Li, Q.: Clip-reid: exploiting vision-language model for image re- identificationwithoutconcretetextlabels.In:ProceedingsoftheAAAIconference on artificial intelligence. vol. 37, pp. 1405–1413 (2023)

2023
[38]

In: Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recogni- tion

Li, T., Liu, J., Zhang, W., Ni, Y., Wang, W., Li, Z.: Uav-human: A large bench- mark for human behavior understanding with unmanned aerial vehicles. In: Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recogni- tion. pp. 16266–16275 (2021)

2021
[39]

In: Proceedings of the IEEE conference on computer vision and pattern recognition

Li, W., Zhao, R., Xiao, T., Wang, X.: Deepreid: Deep filter pairing neural network for person re-identification. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 152–159 (2014)

2014
[40]

Li, Y.J., Chen, Y.C., Lin, Y.Y., Du, X., Wang, Y.C.F.: Recover and identify: A generativedualmodelforcross-resolutionpersonre-identification.In:Proceedings of the IEEE/CVF international conference on computer vision. pp. 8090–8099 (2019)

2019
[41]

In: Proceedings of the IEEE/CVF Winter conference on applications of computer vision

Li, Y.J., Weng, X., Kitani, K.M.: Learning shape representations for person re- identification under clothing change. In: Proceedings of the IEEE/CVF Winter conference on applications of computer vision. pp. 2432–2441 (2021)

2021
[42]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Li, Y., He, J., Zhang, T., Liu, X., Zhang, Y., Wu, F.: Diverse part discovery: Occluded person re-identification with part-aware transformer. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 2898–2907 (2021)

2021
[43]

In: European Conference on Computer Vision

Li, Z., Liu, J., Zhang, Z., Xu, S., Yan, Y.: Cliff: Carrying location information in full frames into human pose and shape estimation. In: European Conference on Computer Vision. pp. 590–606. Springer (2022)

2022
[44]

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

Lin, J., Zeng, A., Wang, H., Zhang, L., Li, Y.: One-stage 3d whole-body mesh recovery with component aware transformer. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 21159–21168 (2023)

2023
[45]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Lin, K., Wang, L., Liu, Z.: End-to-end human pose and mesh reconstruction with transformers. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 1954–1963 (2021)

1954
[46]

In: Proceedings of the IEEE/CVF international conference on computer vision

Lin, K., Wang, L., Liu, Z.: Mesh graphormer. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 12939–12948 (2021)

2021
[47]

IEEE Transactions on Image Processing32, 5075–5086 (2023) Universal Monocular 3D-Enhanced ReID 19

Liu, F., Ye, M., Du, B.: Dual level adaptive weighting for cloth-changing person re-identification. IEEE Transactions on Image Processing32, 5075–5086 (2023) Universal Monocular 3D-Enhanced ReID 19

2023
[48]

In: Proceedings of the 32nd ACM International Conference on Multimedia

Liu, F., Ye, M., Du, B.: Cloth-aware augmentation for cloth-generalized person re-identification. In: Proceedings of the 32nd ACM International Conference on Multimedia. pp. 4053–4062 (2024)

2024
[49]

In: Proceedings of the IEEE/CVF international conference on computer vision

Liu, F., Kim, M., Gu, Z., Jain, A., Liu, X.: Learning clothing and pose invariant 3d shape representation for long-term person re-identification. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 19617–19626 (2023)

2023
[50]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Liu, F., Kim, M., Ren, Z., Liu, X.: Distilling clip with dual guidance for learn- ing discriminative human body shape representation. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 256–266 (2024)

2024
[51]

IEEE Transactions on Mul- timedia23, 4414–4425 (2020)

Liu, H., Tan, X., Zhou, X.: Parameter sharing exploration and hetero-center triplet loss for visible-thermal person re-identification. IEEE Transactions on Mul- timedia23, 4414–4425 (2020)

2020
[52]

ACM Trans

Loper, M., Mahmood, N., Romero, J., Pons-Moll, G., Black, M.J.: SMPL: a skinned multi-person linear model. ACM Trans. Graph.34(6), 248:1–248:16 (2015).https://doi.org/10.1145/2816795.2818013

2015
[53]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops

Luo, H., Gu, Y., Liao, X., Lai, S., Jiang, W.: Bag of tricks and a strong baseline for deep person re-identification. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops. pp. 0–0 (2019)

2019
[54]

In: Proceedings of the IEEE/CVF international conference on computer vision

Miao, J., Wu, Y., Liu, P., Ding, Y., Yang, Y.: Pose-guided feature alignment for occluded person re-identification. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 542–551 (2019)

2019
[55]

In: European Conference on Computer Vision

Moon, G., Lee, K.M.: I2l-meshnet: Image-to-lixel prediction network for accu- rate 3d human pose and mesh estimation from a single rgb image. In: European Conference on Computer Vision. pp. 752–768. Springer (2020)

2020
[56]

v2: Bridging aerial and ground views for person re-identification

Nguyen, H., Nguyen, K., Sridharan, S., Fookes, C.: Ag-reid. v2: Bridging aerial and ground views for person re-identification. IEEE Transactions on Information Forensics and Security19, 2896–2908 (2024)

2024
[57]

In: 2021 9th European Workshop on Visual Information Processing (EUVIP)

Ni, X., Rahtu, E.: Flipreid: closing the gap between training and inference in person re-identification. In: 2021 9th European Workshop on Visual Information Processing (EUVIP). pp. 1–6. IEEE (2021)

2021
[58]

In: 2018 international conference on 3D vision (3DV)

Omran, M., Lassner, C., Pons-Moll, G., Gehler, P., Schiele, B.: Neural body fit- ting: Unifying deep learning and model based human pose and shape estimation. In: 2018 international conference on 3D vision (3DV). pp. 484–494. IEEE (2018)

2018
[59]

In: Proceedings of the Computer Vision and Pattern Recognition Conference

Pang, Z., Wang, J., Zhao, L., Wang, C.: Identity-clothing similarity modeling for unsupervised clothing change person re-identification. In: Proceedings of the Computer Vision and Pattern Recognition Conference. pp. 19251–19260 (2025)

2025
[60]

IEEE signal processing magazine34(6), 96–108 (2017)

Ramachandram, D., Taylor, G.W.: Deep multimodal learning: A survey on recent advances and trends. IEEE signal processing magazine34(6), 96–108 (2017)

2017
[61]

In: Proceedings of the 29th ACM international conference on Multimedia

Rao, H., Hu, X., Cheng, J., Hu, B.: Sm-sge: A self-supervised multi-scale skeleton graph encoding framework for person re-identification. In: Proceedings of the 29th ACM international conference on Multimedia. pp. 1812–1820 (2021)

2021
[62]

In: Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition

Rao, H., Miao, C.: Transg: Transformer-based skeleton graph prototype con- trastive learning with structure-trajectory prompted reconstruction for person re-identification. In: Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition. pp. 22118–22128 (2023)

2023
[63]

In: Proceedings of the IEEE/CVF Conference on Com- puter Vision and Pattern Recognition

Ren, K., Zhang, L.: Implicit discriminative knowledge learning for visible-infrared person re-identification. In: Proceedings of the IEEE/CVF Conference on Com- puter Vision and Pattern Recognition. pp. 393–402 (2024) 20 B. Su et al

2024
[64]

In: 32nd USENIX Security Symposium (USENIX Security 23)

Ren, Y., Wang, Y., Tan, S., Chen, Y., Yang, J.: Person re-identification in 3d space: A{WiFi}vision-based approach. In: 32nd USENIX Security Symposium (USENIX Security 23). pp. 5217–5234 (2023)

2023
[65]

Image and vision computing117, 104335 (2022)

Shi, W., Liu, H., Liu, M.: Iranet: Identity-relevance aware representation for cloth- changing person re-identification. Image and vision computing117, 104335 (2022)

2022
[66]

In: European Conference on Computer Vision

Somers, V., Alahi, A., Vleeschouwer, C.D.: Keypoint promptable re-identification. In: European Conference on Computer Vision. pp. 216–233. Springer (2024)

2024
[67]

In: Proceedings of the IEEE/CVF winter conference on applications of computer vision

Somers,V.,DeVleeschouwer,C.,Alahi,A.:Bodypart-basedrepresentationlearn- ing for occluded person re-identification. In: Proceedings of the IEEE/CVF winter conference on applications of computer vision. pp. 1613–1623 (2023)

2023
[68]

In: Proceedings of the IEEE/CVF conference on computer vision and pat- tern recognition

Sun, X., Zheng, L.: Dissecting person re-identification from the viewpoint of view- point. In: Proceedings of the IEEE/CVF conference on computer vision and pat- tern recognition. pp. 608–617 (2019)

2019
[69]

In: Pro- ceedings of the European conference on computer vision (ECCV)

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

2018
[70]

arXiv preprint arXiv:2408.16684 (2024)

Tan, L., Dai, P., Chen, J., Cao, L., Wu, Y., Ji, R.: Partformer: Awakening latent diverse representation from vision transformer for object re-identification. arXiv preprint arXiv:2408.16684 (2024)

2024 arXiv
[71]

In: Proceedings of the 26th ACM international conference on Multimedia

Wang, G., Yuan, Y., Chen, X., Li, J., Zhou, X.: Learning discriminative features with multiple granularities for person re-identification. In: Proceedings of the 26th ACM international conference on Multimedia. pp. 274–282 (2018)

2018
[72]

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

Wang, S., Wang, Y., Wu, R., Jiao, B., Wang, W., Wang, P.: Secap: self-calibrating and adaptive prompts for cross-view person re-identification in aerial-ground net- works. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 22119–22128 (2025)

2025
[73]

In: Proceedings of the AAAI conference on artificial intelligence

Wang, T., Liu, H., Song, P., Guo, T., Shi, W.: Pose-guided feature disentangling for occluded person re-identification based on transformer. In: Proceedings of the AAAI conference on artificial intelligence. vol. 36, pp. 2540–2549 (2022)

2022
[74]

In: Proceedings of the 28th ACM international conference on multimedia

Wang, Y., Liao, S., Shao, L.: Surpassing real-world source training data: Random 3d characters for generalizable person re-identification. In: Proceedings of the 28th ACM international conference on multimedia. pp. 3422–3430 (2020)

2020
[75]

In: Proceedings of the IEEE conference on computer vision and pattern recognition

Wang, Y., Chen, Z., Wu, F., Wang, G.: Person re-identification with cascaded pairwise convolutions. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 1470–1478 (2018)

2018
[76]

In: Proceedings of the 31st ACM International Conference on Multimedia

Wang, Y., Yu, H., Yan, Y., Song, S., Liu, B., Lu, Y.: Exploring shape embed- ding for cloth-changing person re-identification via 2d-3d correspondences. In: Proceedings of the 31st ACM International Conference on Multimedia. pp. 7121– 7130 (2023)

2023
[77]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Wang, Z., Zhu, F., Tang, S., Zhao, R., He, L., Song, J.: Feature erasing and diffusion network for occluded person re-identification. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 4754– 4763 (2022)

2022
[78]

In: Proceedings of the IEEE conference on computer vision and pattern recognition

Wei, L., Zhang, S., Gao, W., Tian, Q.: Person transfer gan to bridge domain gap for person re-identification. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 79–88 (2018)

2018
[79]

In: Proceedings of the IEEE international conference on computer vision

Wu, A., Zheng, W.S., Yu, H.X., Gong, S., Lai, J.: Rgb-infrared cross-modality person re-identification. In: Proceedings of the IEEE international conference on computer vision. pp. 5380–5389 (2017) Universal Monocular 3D-Enhanced ReID 21

2017
[80]

IEEE transactions on neural networks and learning systems30(11), 3347–3359 (2019)

Wu, L., Wang, Y., Shao, L., Wang, M.: 3-d personvlad: Learning deep global rep- resentations for video-based person reidentification. IEEE transactions on neural networks and learning systems30(11), 3347–3359 (2019)

2019
[81]

arXiv preprint arXiv:2507.00506 (2025)

Xie, Y., Cheng, Y., Wu, J., Zhang, H., Zhou, Y., Han, S.: Scing: Towards more efficient and robust person re-identification through selective cross-modal prompt tuning. arXiv preprint arXiv:2507.00506 (2025)

2025 arXiv
[82]

Sci- entific Reports15(1), 34912 (2025)

Xu, H., Gao, R.: A dual-branch pedestrian re-identification method cphmnet based on multi-dimensional feature fusion and integrated pose estimation. Sci- entific Reports15(1), 34912 (2025)

2025
[83]

In: IJCAI

Xu, W., Liu, H., Shi, W., Miao, Z., Lu, Z., Chen, F.: Adversarial feature disentan- glement for long-term person re-identification. In: IJCAI. pp. 1201–1207 (2021)

2021
[84]

In: Proceedings of the IEEE/CVF international conference on computer vision

Yan, C., Pang, G., Jiao, J., Bai, X., Feng, X., Shen, C.: Occluded person re- identification with single-scale global representations. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 11875–11884 (2021)

2021
[85]

IEEE transactions on pattern analysis and machine intelligence43(6), 2029–2046 (2019)

Yang, Q., Wu, A., Zheng, W.S.: Person re-identification by contour sketch under moderate clothing change. IEEE transactions on pattern analysis and machine intelligence43(6), 2029–2046 (2019)
[86]

In: Proceedings of the IEEE/CVFconferenceoncomputervisionandpatternrecognition.pp.1472–1481 (2023)

Yang, Z., Lin, M., Zhong, X., Wu, Y., Wang, Z.: Good is bad: Causality inspired cloth-debiasing for cloth-changing person re-identification. In: Proceedings of the IEEE/CVFconferenceoncomputervisionandpatternrecognition.pp.1472–1481 (2023)

2023
[87]

IEEE transactions on pattern analysis and machine intelligence44(6), 2872–2893 (2021)

Ye, M., Shen, J., Lin, G., Xiang, T., Shao, L., Hoi, S.C.: Deep learning for person re-identification: A survey and outlook. IEEE transactions on pattern analysis and machine intelligence44(6), 2872–2893 (2021)

2021
[88]

IEEE Transactions on Information Forensics and Security16, 728–739 (2020)

Ye, M., Shen, J., Shao, L.: Visible-infrared person re-identification via homoge- neous augmented tri-modal learning. IEEE Transactions on Information Forensics and Security16, 728–739 (2020)

2020
[89]

Knowledge- Based Systems283, 111200 (2024)

Yu, Z., Tiwari, P., Hou, L., Li, L., Li, W., Jiang, L., Ning, X.: Mv-reid: 3d multi- view transformation network for occluded person re-identification. Knowledge- Based Systems283, 111200 (2024)

2024
[90]

Image and Vision Computing116, 104330 (2021)

Zang, X., Li, G., Gao, W., Shu, X.: Learning to disentangle scenes for person re-identification. Image and Vision Computing116, 104330 (2021)

2021
[91]

IEEE Trans- actions on Image processing30, 8913–8925 (2021)

Zhang, G., Ge, Y., Dong, Z., Wang, H., Zheng, Y., Chen, S.: Deep high-resolution representation learning for cross-resolution person re-identification. IEEE Trans- actions on Image processing30, 8913–8925 (2021)

2021
[92]

In: Proceedings of the IEEE/CVF international conference on computer vision

Zhang, H., Tian, Y., Zhou, X., Ouyang, W., Liu, Y., Wang, L., Sun, Z.: Pymaf: 3d human pose and shape regression with pyramidal mesh alignment feedback loop. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 11446–11456 (2021)

2021
[93]

Computers14(11), 502 (2025)

Zhang, J., Wu, Y., Jiang, H.: Survey on monocular metric depth estimation. Computers14(11), 502 (2025)

2025
[94]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Zhang, T., Xie, L., Wei, L., Zhuang, Z., Zhang, Y., Li, B., Tian, Q.: Unrealperson: An adaptive pipeline towards costless person re-identification. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 11506–11515 (2021)

2021
[95]

Artificial Intelligence Review58(2), 53 (2024)

Zhao, C., Qin, Y., Zhang, B., Zhao, Y., Wu, B.: An end-to-end occluded person re-identification network with smoothing corrupted feature prediction. Artificial Intelligence Review58(2), 53 (2024)

2024
[96]

IEEE transactions on image processing28(9), 4500–4509 (2019) 22 B

Zheng, L., Huang, Y., Lu, H., Yang, Y.: Pose-invariant embedding for deep person re-identification. IEEE transactions on image processing28(9), 4500–4509 (2019) 22 B. Su et al

2019
[97]

In: Proceedings of the IEEE international conference on computer vision

Zheng, L., Shen, L., Tian, L., Wang, S., Wang, J., Tian, Q.: Scalable person re- identification: A benchmark. In: Proceedings of the IEEE international conference on computer vision. pp. 1116–1124 (2015)

2015
[98]

arXiv preprint arXiv:1610.02984 (2016)

Zheng, L., Yang, Y., Hauptmann, A.G.: Person re-identification: Past, present and future. arXiv preprint arXiv:1610.02984 (2016)

2016 arXiv
[99]

International Journal of Computer Vision130(1), 136–156 (2022)

Zheng, W.S., Hong, J., Jiao, J., Wu, A., Zhu, X., Gong, S., Qin, J., Lai, J.: Joint bilateral-resolution identity modeling for cross-resolution person re-identification. International Journal of Computer Vision130(1), 136–156 (2022)

2022
[100]

IEEE Transactions on Pattern Analysis and Machine Intelligence 47(3), 1362–1380 (2024)

Zheng, W.S., Yan, J., Peng, Y.X.: A versatile framework for multi-scene person re- identification. IEEE Transactions on Pattern Analysis and Machine Intelligence 47(3), 1362–1380 (2024)

2024
[101]

IEEE Transactions on Neural Networks and Learn- ing Systems35(6), 7534–7547 (2022)

Zheng, Z., Wang, X., Zheng, N., Yang, Y.: Parameter-efficient person re- identification in the 3d space. IEEE Transactions on Neural Networks and Learn- ing Systems35(6), 7534–7547 (2022)

2022
[102]

In: Proceedings of the IEEE/CVF international conference on computer vision

Zhou, K., Yang, Y., Cavallaro, A., Xiang, T.: Omni-scale feature learning for per- son re-identification. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 3702–3712 (2019)

2019
[103]

In: European conference on computer vision

Zhu, K., Guo, H., Liu, Z., Tang, M., Wang, J.: Identity-guided human semantic parsing for person re-identification. In: European conference on computer vision. pp. 346–363. Springer (2020)

2020
[104]

Zhu, K., Guo, H., Zhang, S., Wang, Y., Liu, J., Wang, J., Tang, M.: Aaformer: Auto-alignedtransformerforpersonre-identification.IEEEtransactionsonneural networks and learning systems35(12), 17307–17317 (2023)

2023
[105]

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

Zhu, Y., Li, A., Tang, Y., Zhao, W., Zhou, J., Lu, J.: Dpmesh: Exploiting diffu- sion prior for occluded human mesh recovery. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 1101–1110 (2024)

2024
[106]

Neurocomputing386, 97–109 (2020)

Zhu, Y., Yang, Z., Wang, L., Zhao, S., Hu, X., Tao, D.: Hetero-center loss for cross-modality person re-identification. Neurocomputing386, 97–109 (2020)

2020

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.

[1] [1]

In: European conference on computer vision

Bogo, F., Kanazawa, A., Lassner, C., Gehler, P., Romero, J., Black, M.J.: Keep it smpl: Automatic estimation of 3d human pose and shape from a single image. In: European conference on computer vision. pp. 561–578. Springer (2016)

2016

[2] [2]

Knowledge-Based Systems307, 112737 (2025)

Che, Q.H., Nguyen, L.C., Luu, D.T., Nguyen, V.T.: Enhancing person re- identification via uncertainty feature fusion method and auto-weighted measure combination. Knowledge-Based Systems307, 112737 (2025)

2025

[3] [3]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Chen, C., Ye, M., Jiang, D.: Towards modality-agnostic person re-identification with descriptive query. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 15128–15137 (2023)

2023

[4] [4]

In: Proceed- ings of the IEEE/CVF conference on computer vision and pattern recognition

Chen, J., Jiang, X., Wang, F., Zhang, J., Zheng, F., Sun, X., Zheng, W.S.: Learn- ing 3d shape feature for texture-insensitive person re-identification. In: Proceed- ings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 8146–8155 (2021)

2021

[5] [5]

In: Proceedings of the IEEE/CVF international conference on computer vision

Chen, T., Ding, S., Xie, J., Yuan, Y., Chen, W., Yang, Y., Ren, Z., Wang, Z.: Abd-net: Attentive but diverse person re-identification. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 8351–8361 (2019)

2019

[6] [6]

In: Proceedings of the IEEE/CVF conference on com- puter vision and pattern recognition

Chen, W., Xu, X., Jia, J., Luo, H., Wang, Y., Wang, F., Jin, R., Sun, X.: Be- yond appearance: a semantic controllable self-supervised learning framework for human-centric visual tasks. In: Proceedings of the IEEE/CVF conference on com- puter vision and pattern recognition. pp. 15050–15061 (2023)

2023

[7] [7]

In: 2024 International Joint Conference on Neural Net- works (IJCNN)

Chen, Z., Ge, Y.: Part-attention based model make occluded person re- identification stronger. In: 2024 International Joint Conference on Neural Net- works (IJCNN). pp. 1–8. IEEE (2024)

2024

[8] [8]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Cheng, Z., Dong, Q., Gong, S., Zhu, X.: Inter-task association critic for cross- resolution person re-identification. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 2605–2615 (2020)

2020