Mixed High-Order Attention Network for Person Re-Identification

Binghui Chen; Jiani Hu; Weihong Deng

REVIEW 3 major objections 4 minor 67 references

Mixed High-Order Attention Network for Person Re-Identification

T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read The paper proposes that person re-identification improves when attention is built from high-order statistics of convolutional features, and it reports state-of-the-art results on three benchmarks using a mixed high-order attention network.

desk verdict The HOA high-order attention idea is real and the experiments are consistent, but the adversarial order-diversity objective in Eq. 10 is degenerate as written, so the paper's main theoretical justification needs revision. read the letter →

arxiv 1908.05819 v1 pith:B6IBEYOQ submitted 2019-08-16 cs.CV

Binghui Chen , Weihong Deng , Jiani Hu This is my paper · ORCID

classification cs.CV

keywords personre-identificationhigh-orderattentionmechanismzero-shotlearningadversarialconvolutionalneuralnetworkspedestrianretrieval

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

Person re-identification models usually guide a network with spatial or channel attention masks built from first-order statistics. This paper tries to establish that attention maps built from high-order statistics of convolutional activations are more discriminative, and that mixing several orders explicitly is the right way to train them. To that end it introduces the High-Order Attention (HOA) module and the Mixed High-Order Attention Network (MHN), which runs HOA modules of orders 1 through k and uses an adversarial loss to keep the branches at distinct orders. If the claim is right, attention in re-identification no longer needs hand-designed coarse masks, and the reported experiments support it with new state-of-the-art results on Market-1501, DukeMTMC-ReID, and CUHK03-NP, including 95.1% rank-1 accuracy and 85.0% mAP on Market-1501.

What carries the argument

The central object is the High-Order Attention (HOA) module, a 1x1-convolution implementation of a high-order polynomial predictor that outputs a reweighting map $Y = A(X) \odot X$. The mixed design is the Mixed High-Order Attention Network (MHN), which places HOA modules of orders $R=1,2,\dots,k$ between an early encoder and a shared later encoder, so that diverse orders contribute to the embedding while parameter growth stays small. The third piece is the adversarial order-diversity loss $L_{adv} = \max_{HOA} \min_F \sum_{j\neq j'} \|F(f_j) - F(f_{j'})\|_2^2$, which the paper uses to keep the branches from converging to the same effective order.

What would settle it

Train MHN-6 with and without the adversarial loss while logging the norm of each higher-order term $\hat{\alpha}_r$ in every HOA branch and the pairwise distances between stream features before $F$. If removing $L_{adv}$ leaves the higher-order terms just as active, or if adding it makes $F$ output near-constant vectors while the branches stay collapsed, then the claimed mechanism is not what drives the accuracy gains.

Watch

Extended reading notes

Core claim

The core discovery is that high-order attention—attention computed from products of multiple projections of a local descriptor—captures subtle pedestrian differences that first-order spatial and channel attention miss. Concretely, HOA evaluates a per-location polynomial predictor $a(x)=\sum_{r=1}^R \hat{\alpha}_r^T z_r$, with $z_1 = \hat{v}^T x$ and $z_r$ formed by the elementwise product of $r$ projected copies of $x$, applies ReLU and sigmoid, and uses the resulting map to reweight the feature tensor. MHN assembles $k$ such modules with different orders, shares the later backbone layers across streams so the extra cost stays modest, and concatenates the stream features at test time. An adversarial objective over the HOA branches and an encoding network $F$ is added to stop higher-order branches from collapsing to lower-order behavior. The paper reports that MHN-6 improves both IDE and PCB baselines by clear margins and reaches the best published numbers on all three benchmark datasets.

Load-bearing premise

The load-bearing premise is that the adversarial loss in Eq. 10 really prevents order collapse—that the min-max game forces the HOA branches to use genuinely different orders rather than the adversary trivially mapping all features to the same constant while the branches still collapse.

Editorial extensions

If this is right

Adding HOA modules of increasing orders produces steady gains over both IDE and PCB baselines on Market-1501, DukeMTMC-ReID, and CUHK03-NP, with the largest gains coming from moving from order 2 to order 6.
The adversarial constraint contributes beyond simply mixing orders: MHN-6 with $L_{adv}$ beats MHN-6 without it on both Market-1501 and DukeMTMC-ReID, supporting the order-collapse story.
The method is model-agnostic in the sense that the same MHN wrapper improves both IDE and PCB, and parameter counts stay comparable to the baselines at about 26.8 million parameters for MHN-6 on IDE versus 24.2 million for IDE.
At test time, L2-normalized features from all HOA branches are concatenated, and this single representation is used for ranking without re-ranking tricks.

Reading between the lines

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

A direct test of the mechanism would be to measure the effective order of each HOA branch during training; if the branches remain distinct even without $L_{adv}$, the adversarial loss is not doing the work the paper assigns to it.
The HOA module is not person-specific, so the same high-order attention recipe could plausibly transfer to other fine-grained recognition tasks such as vehicle or face re-identification, where viewpoint changes also create subtle part-level differences.
The zero-shot framing suggests that any re-identification model is vulnerable to biased learning; one could audit other attention methods by checking whether their attention maps change when evaluated on unseen identities rather than training identities.
Because the reported gains are before re-ranking, combining MHN features with k-reciprocal re-ranking may lift the numbers further, though the paper does not claim this.

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.

Desk Editor's Note

read the letter

Quick take: the core idea, high-order attention via factorized polynomial predictors, is a legitimate and useful twist on high-order pooling, and the experiments back it up. But the adversarial diversity loss in Eq. 10 is degenerate as written—unconstrained F can collapse to a constant and make min_F Ladv = 0—so the paper's explanation for why that loss works is not valid. The empirical result still stands, but the theory is unsupported.

The HOA module is the best part. Reformulating attention map generation as a polynomial predictor with rank-1 tensor decomposition is clean, and the claim that channel attention is a special case with R=1 is credible. Applying the module between two ResNet stages is model-agnostic, and the ablations (varying k, insertion position, nonlinearity, with vs. without adversarial loss) are more thorough than most ReID papers. The re-implementations of IDE and PCB under the same training config are good practice. Across Market-1501, DukeMTMC-ReID, and CUHK03-NP, the gains over re-implemented baselines are consistent and sometimes large, so the main empirical claim appears solid.

The soft spot is exactly where your concern lands. Eq. 10 is a max-min problem over HOA modules and F, with F as two FC layers. A constant mapping F(x)=c is admissible, giving zero pairwise distance for any HOA configuration. Thus max_HOA min_F Ladv = 0, and the claimed Nash equilibrium argument—that the only way to make Ladv large is to diversify HOA orders—is wrong. The paper never analyzes the actual training dynamics. It's possible the adversarial training helps empirically through some implicit regularization or finite-time effect, and Table 5 suggests it does, but the stated rationale is not supported. Anyone citing this paper for the adversarial diversity mechanism should be careful.

Missing items: the supplementary derivation for Eq. 5 to Eq. 6 is referenced but not provided in the arXiv version; that's a minor accessibility issue. No error bars or multiple seeds, which is typical in this literature but still limits how strongly we can read a 2-point gain. The code link is unverified.

Citation pattern: self-citations [5,6] are used to motivate the zero-shot bias premise; they are the original source of that idea, so it's not inappropriate.

Bottom line: this is a paper worth a serious referee, but the referee should require the authors to either fix or significantly temper the adversarial claim. The HOA module and the empirical study are enough to justify publication after revision. I'd bring it to a reading group to discuss the min-max degeneracy.

Referee Report

3 major / 4 minor

Summary. The paper proposes a High-Order Attention (HOA) module that computes attention maps from high-order statistics of local descriptors via a polynomial predictor with rank-1 tensor decomposition, and a Mixed High-Order Attention Network (MHN) that combines HOA modules of different orders (R=1..k) for person re-identification. The method is motivated by framing ReID as zero-shot learning, where deep models may exhibit biased learning, and an adversarial constraint is introduced to prevent order collapse among HOA modules. On Market-1501, DukeMTMC-ReID, and CUHK03-NP, MHN-6 improves over the authors' re-implemented IDE and PCB baselines, e.g., 95.1% rank-1 and 85.0% mAP vs. PCB's 93.1% and 78.6% on Market-1501. The paper includes ablations on the number of HOA modules, the adversarial term, the nonlinearity, the insertion position, and model size.

Significance. If the claims hold, the paper offers a modular and model-agnostic attention mechanism, a diversity-promoting training objective, and competitive results on three standard benchmarks. The writing is clear, the experimental protocol is described in reasonable detail, and the code URL is provided. The ablation tables consistently support the empirical value of adding HOA modules to both IDE and PCB. However, the theoretical justification of the adversarial order-diversity term in Eq. (10) is invalid as written, which weakens the third contribution and the explanation of why MHN works, even though it does not by itself overturn the reported accuracy numbers.

major comments (3)

[§3.3, Eq. (10)] The adversarial order-diversity objective is degenerate as written. Since F is implemented as two fully-connected layers, a constant map F(f)=c for all f is feasible (e.g., zero weights and equal biases), so ||F(f_j)-F(f_j')||_2^2=0 for every pair and min_F L_adv=0 for every HOA configuration. Consequently max_{HOA} min_F L_adv = 0, and the claim that 'the only solution is to make the HOA modules have different orders' is not correct. The stated mechanism for preventing order collapse is therefore unsupported without additional restrictions on F (such as normalizing the output statistics, spectral normalization, or a margin-based formulation) and without an analysis of the finite-time training dynamics. The empirical gains in Table 5 are not directly falsified, but they cannot be attributed to the stated Nash-equilibrium argument in the current form.
[§3.2, Eqs. (5)-(6)] The reparameterization step from Eq. (5) to Eq. (6) is not derived in the paper; the text says 'simple proof is in Supplementary file', but the arXiv version contains no supplementary material. This step is load-bearing because it justifies the efficient implementation of the HOA module and the reduction of the parameter set into the matrices w_hat_1 and alpha_hat_r. Please include the proof in the main text or make the supplementary file available, and state explicitly which parameters are absorbed into which new matrices.
[§3.3 and §4.2, Table 5] The paper does not specify how the min-max objective in Eq. (11) is optimized (e.g., alternating updates, number of discriminator updates per generator update, gradient reversal, or learning-rate schedules), so the adversarial training is not reproducible from the text. More importantly, no diagnostic is provided to show that HOA modules actually learn different orders. I recommend reporting a measure of effective order for each branch (e.g., the norm of the higher-order terms or the gradient of L_adv with respect to each branch) with and without L_adv, together with a pairwise diversity metric on the features f_j before concatenation. This would establish whether the gains in Table 5 come from order diversity rather than from the extra regularizer acting as a generic penalty.

minor comments (4)

[§3.3] The text says P1 ranges 'from conv1 to layer22'; this appears to be a typo for 'layer2', since Table 8 uses 'layer2' and the ablation compares layer1, layer2, and layer3.
[§4, Implementation] The sentence 'we set all Dr|R r=1 to be 64' is grammatically unclear; it should read something like 'we set D_r=64 for all r=1,...,R'.
[§3.2, Eq. (8)] Applying 'sigmoid' to a vector-valued a(x) should be described as an element-wise operation to avoid ambiguity.
[§4.2, Table 4] In Table 4, PCB*+era degrades relative to PCB* (e.g., Market-1501 rank-1 drops from 93.1 to 91.9), which is inconsistent with the usual behavior of random erasing; since random erasing is applied to all MHN runs, this should be discussed or clarified.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the MHN results are externally validated on standard benchmarks, and the self-citations [5,6] are motivational rather than load-bearing reductions.

full rationale

The paper's central claims are validated against external ReID benchmarks (Market-1501, DukeMTMC-ReID, CUHK03-NP) with standard rank-1/mAP metrics, and no reported number is a fitted input renamed as a prediction. The HOA module is derived from a linear polynomial predictor using tensor decomposition, and the claim that a k-th order HOA contains lower-order sub-terms follows analytically from Eq. 7, not from the authors' prior work. The citations [5,6] are used only to motivate the zero-shot 'biased learning behavior' premise; they are not invoked as a uniqueness theorem, do not forbid alternative designs, and are not the source of the empirical results. The statement that channel attention is a special case of HOA with R=1 is a special-case observation, not a circular derivation of MHN's performance. The adversarial objective in Eq. 10 has a genuine theoretical gap (the inner minimization over F could be trivially satisfied by a constant mapping, so the claimed order-diversity guarantee is not established), but this is an unsupported correctness claim, not a circularity: the loss is not defined in terms of the benchmark outcomes it is said to explain, and the ablation in Table 5 is an empirical measurement rather than a construction. Accordingly, no step in the derivation reduces to its own inputs, and the paper merits a circularity score of 0.

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

The method's central operation is a factorized high-order polynomial predictor over local descriptors. It relies on rank-1 tensor decomposition and a low-rank approximation of the first-order weight matrix. The zero-shot biased-learning premise is cited from the authors' own prior work. The adversarial diversity mechanism is the most fragile assumption because it is not formally justified.

free parameters (4)

k, number of HOA modules / maximum order = 6 (MHN-6; 2 and 4 evaluated; k=8 not reported)
Performance improves with k up to 6, then gains are said to be few. This is an architecture choice tuned on the target datasets.
D_r, number of rank-1 terms per order = 64
Set to 64 for all orders r; a capacity hyperparameter chosen without reported sensitivity analysis.
lambda, weight of adversarial loss = not reported
Coefficient in Eq. 11; no value or sensitivity study is given.
HOA insertion position (P1/P2 split) = after layer2
Selected from the ablation in Table 8 among three possible splits.

assumptions (4)

standard math The rank-r weight tensor w_r can be approximated by D_r rank-1 tensors via tensor decomposition.
Equation 3; standard low-rank tensor approximation, but it introduces an approximation error whose effect is not analyzed.
domain assumption The first-order weight matrix \hat w1 can be factored as \hat v \hat alpha1 with D1 << C.
Introduced between Eq. 6 and Eq. 7; a low-rank assumption without stated justification or validation.
domain assumption Deep models in zero-shot settings exhibit biased and partial learning that causes HOA modules to collapse to lower orders.
Sec. 3.3; cited from the authors' prior work [5,6] and used to motivate the adversarial loss.
ad hoc to paper Maximizing pairwise feature discrepancies under the Eq. 10 game forces HOA branches to have distinct orders.
Sec. 3.3; no formal proof is given, and the min over F can be driven to zero by a constant mapping, which would undermine the claimed mechanism.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Mixed High-Order Attention Network for Person Re-Identification." pith.science (2026). https://pith.science/paper/B6IBEYOQ

@misc{pith2026190805819,
  author       = {Pith},
  title        = {Pith review of: Mixed High-Order Attention Network for Person Re-Identification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/B6IBEYOQ}},
  note         = {Machine review of arXiv:1908.05819}
}

read the original abstract

Attention has become more attractive in person reidentification (ReID) as it is capable of biasing the allocation of available resources towards the most informative parts of an input signal. However, state-of-the-art works concentrate only on coarse or first-order attention design, e.g. spatial and channels attention, while rarely exploring higher-order attention mechanism. We take a step towards addressing this problem. In this paper, we first propose the High-Order Attention (HOA) module to model and utilize the complex and high-order statistics information in attention mechanism, so as to capture the subtle differences among pedestrians and to produce the discriminative attention proposals. Then, rethinking person ReID as a zero-shot learning problem, we propose the Mixed High-Order Attention Network (MHN) to further enhance the discrimination and richness of attention knowledge in an explicit manner. Extensive experiments have been conducted to validate the superiority of our MHN for person ReID over a wide variety of state-of-the-art methods on three large-scale datasets, including Market-1501, DukeMTMC-ReID and CUHK03-NP. Code is available at http://www.bhchen.cn/.

Figures

Figures reproduced from arXiv: 1908.05819 by the authors.

**Figure 1.** Attention comparison. (1) Spatial attention uses softmax-like gated functions to produce a spatial mask. (2) Channel attention [19] uses global average pooling and fully connected layers to produce a scale vector. (3) Our high-order attention uses high-order polynomial predictor to produce scale maps that contain high-order statistics of convolutional activations. tion of feature representations of pedestrian imag… view at source ↗

**Figure 2.** Illustration of High-Order Attention (HOA) modules. tensor X , we generalize Eq.9. Specifically, we share the learnable weights in A(x) among different spatial locations of X and let A(X ) = {A(x(1,1)), · · · , A(x(H,W))}, where x(h,w) indicates a local descriptor at spatial location point (h, w) of X . Employing this attention map A(X ) in CNN has two benefits. (1) sharing weights among different spatial locations… view at source ↗

**Figure 3.** Illustration of Mixed High-Order Attention Network (MHN). Our MHN is model-agnostic, it can be applied in both IDE [63] and PCB [43] architectures, here for clarity, we take ResNet50 [18] based IDE for example. The adversary constraint is used to regularize the order of HOA modules. channel-attention [19] are equivalent to our HOA module when R = 1 (regardless of the spatial dimensions and see Fig.2.(a)). In summary… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

67 extracted references · 55 canonical work pages

[1]

https://github.com/layumi/Person_reID_ baseline_pytorch. 6

work page
[2]

Higher-order in- tegration of hierarchical convolutional activations for ﬁne- grained visual categorization

Sijia Cai, Wangmeng Zuo, and Lei Zhang. Higher-order in- tegration of hierarchical convolutional activations for ﬁne- grained visual categorization. In Proceedings of the IEEE International Conference on Computer Vision , pages 511– 520, 2017. 3

work page 2017
[3]

Multi-level factorisation net for person re-identiﬁcation

Xiaobin Chang, Timothy M Hospedales, and Tao Xiang. Multi-level factorisation net for person re-identiﬁcation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 2109–2118, 2018. 6, 7

work page 2018
[4]

Synthesized classiﬁers for zero-shot learning

Soravit Changpinyo, Wei-Lun Chao, Boqing Gong, and Fei Sha. Synthesized classiﬁers for zero-shot learning. In Pro- ceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 5327–5336, 2016. 3

work page 2016
[5]

Energy confused adver- sarial metric learning for zero-shot image retrieval and clus- tering

Binghui Chen and Weihong Deng. Energy confused adver- sarial metric learning for zero-shot image retrieval and clus- tering. In AAAI Conference on Artiﬁcial Intelligence, 2019. 2, 3, 5

work page 2019
[6]

Hybrid-attention based decoupled metric learning for zero-shot image retrieval

Binghui Chen and Weihong Deng. Hybrid-attention based decoupled metric learning for zero-shot image retrieval. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 2750–2759, 2019. 3

work page 2019
[7]

Noisy soft- max: Improving the generalization ability of dcnn via post- poning the early softmax saturation

Binghui Chen, Weihong Deng, and Junping Du. Noisy soft- max: Improving the generalization ability of dcnn via post- poning the early softmax saturation. In The IEEE Confer- ence on Computer Vision and Pattern Recognition (CVPR) , July 2017. 2

work page 2017
[8]

Vir- tual class enhanced discriminative embedding learning

Binghui Chen, Weihong Deng, and Haifeng Shen. Vir- tual class enhanced discriminative embedding learning. In Advances in Neural Information Processing Systems , pages 1946–1956, 2018. 2

work page 1946

Show all 67 references

[9]

Group consistent similarity learning via deep crf for person re-identiﬁcation

Dapeng Chen, Dan Xu, Hongsheng Li, Nicu Sebe, and Xiao- gang Wang. Group consistent similarity learning via deep crf for person re-identiﬁcation. In Proceedings of the IEEE Con- ference on Computer Vision and Pattern Recognition, pages 8649–8658, 2018. 6, 7

2018
[10]

Sca-cnn: Spatial and channel-wise attention in convolutional networks for image captioning

Long Chen, Hanwang Zhang, Jun Xiao, Liqiang Nie, Jian Shao, Wei Liu, and Tat-Seng Chua. Sca-cnn: Spatial and channel-wise attention in convolutional networks for image captioning. In Proceedings of the IEEE Conference on Com- puter Vision and Pattern Recognition , pages 5659–5667,
[11]

Deep ﬁlter banks for texture recognition and segmentation

Mircea Cimpoi, Subhransu Maji, and Andrea Vedaldi. Deep ﬁlter banks for texture recognition and segmentation. InPro- ceedings of the IEEE conference on computer vision and pat- tern recognition, pages 3828–3836, 2015. 4

2015
[12]

Kernel pooling for convolutional neu- ral networks

Yin Cui, Feng Zhou, Jiang Wang, Xiao Liu, Yuanqing Lin, and Serge Belongie. Kernel pooling for convolutional neu- ral networks. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 2921–2930,
[13]

Zero-shot video retrieval using content and concepts

Jeffrey Dalton, James Allan, and Pranav Mirajkar. Zero-shot video retrieval using content and concepts. In Proceedings of the 22nd ACM international conference on Information & Knowledge Management, pages 1857–1860. ACM, 2013. 3

2013
[14]

Multimodal com- pact bilinear pooling for visual question answering and vi- sual grounding

Akira Fukui, Dong Huk Park, Daylen Yang, Anna Rohrbach, Trevor Darrell, and Marcus Rohrbach. Multimodal com- pact bilinear pooling for visual question answering and vi- sual grounding. In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing , ...

2016
[15]

Compact bilinear pooling

Yang Gao, Oscar Beijbom, Ning Zhang, and Trevor Darrell. Compact bilinear pooling. In Proceedings of the IEEE con- ference on computer vision and pattern recognition , pages 317–326, 2016. 3

2016
[16]

Person re-identiﬁcation ranking optimisation by discriminant context information analysis

Jorge Garcia, Niki Martinel, Christian Micheloni, and Al- fredo Gardel. Person re-identiﬁcation ranking optimisation by discriminant context information analysis. InProceedings of the IEEE International Conference on Computer Vision , pages 1305–1313, 2015. 2

2015
[17]

Generative adversarial nets

Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Generative adversarial nets. In Advances in neural information processing systems, pages 2672–2680,
[18]

Deep residual learning for image recognition

Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proceed- ings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016. 5

2016
[19]

Squeeze-and-excitation net- works

Jie Hu, Li Shen, and Gang Sun. Squeeze-and-excitation net- works. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 7132–7141, 2018. 1, 2, 3, 4, 5, 8

2018
[20]

Spatial transformer networks

Max Jaderberg, Karen Simonyan, Andrew Zisserman, et al. Spatial transformer networks. In Advances in neural infor- mation processing systems, pages 2017–2025, 2015. 2

2017
[21]

Human seman- tic parsing for person re-identiﬁcation

Mahdi M Kalayeh, Emrah Basaran, Muhittin G ¨okmen, Mustafa E Kamasak, and Mubarak Shah. Human seman- tic parsing for person re-identiﬁcation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recogni- tion, pages 1062–1071, 2018. 1

2018
[22]

Hadamard product for low-rank bilinear pooling

Jin-Hwa Kim, Kyoung-Woon On, Woosang Lim, Jeonghee Kim, Jung-Woo Ha, and Byoung-Tak Zhang. Hadamard product for low-rank bilinear pooling. arXiv preprint arXiv:1610.04325, 2016. 3

2016 arXiv
[23]

Tensor decompositions and applications

Tamara G Kolda and Brett W Bader. Tensor decompositions and applications. SIAM review, 51(3):455–500, 2009. 3

2009
[24]

Learning deep context-aware features over body and latent parts for person re-identiﬁcation

Dangwei Li, Xiaotang Chen, Zhang Zhang, and Kaiqi Huang. Learning deep context-aware features over body and latent parts for person re-identiﬁcation. InProceedings of the IEEE Conference on Computer Vision and Pattern Recogni- tion, pages 384–393, 2017. 2

2017
[25]

Diversity regularized spatiotemporal attention for video- based person re-identiﬁcation

Shuang Li, Slawomir Bak, Peter Carr, and Xiaogang Wang. Diversity regularized spatiotemporal attention for video- based person re-identiﬁcation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , pages 369–378, 2018. 1, 2, 3, 8

2018
[26]

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

Wei Li, Rui Zhao, Tong Xiao, and Xiaogang Wang. Deep- reid: Deep ﬁlter pairing neural network for person re- identiﬁcation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , pages 152–159,
[27]

Harmonious at- tention network for person re-identiﬁcation

Wei Li, Xiatian Zhu, and Shaogang Gong. Harmonious at- tention network for person re-identiﬁcation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 2285–2294, 2018. 1, 2, 3, 4, 6, 7, 8

2018
[28]

Dis- criminative learning of latent features for zero-shot recogni- tion

Yan Li, Junge Zhang, Jianguo Zhang, and Kaiqi Huang. Dis- criminative learning of latent features for zero-shot recogni- tion. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 7463–7471, 2018. 3

2018
[29]

Bilinear cnn models for ﬁne-grained visual recognition

Tsung-Yu Lin, Aruni RoyChowdhury, and Subhransu Maji. Bilinear cnn models for ﬁne-grained visual recognition. In Proceedings of the IEEE international conference on com- puter vision, pages 1449–1457, 2015. 2

2015
[30]

Hydraplus-net: Attentive deep features for pedestrian analysis

Xihui Liu, Haiyu Zhao, Maoqing Tian, Lu Sheng, Jing Shao, Shuai Yi, Junjie Yan, and Xiaogang Wang. Hydraplus-net: Attentive deep features for pedestrian analysis. In Proceed- ings of the IEEE international conference on computer vi- sion, pages 350–359, 2017. 1

2017
[31]

Multi- camera activity correlation analysis

Chen Change Loy, Tao Xiang, and Shaogang Gong. Multi- camera activity correlation analysis. In 2009 IEEE Con- ference on Computer Vision and Pattern Recognition, pages 1988–1995. IEEE, 2009. 1

2009
[32]

Time- delayed correlation analysis for multi-camera activity un- derstanding

Chen Change Loy, Tao Xiang, and Shaogang Gong. Time- delayed correlation analysis for multi-camera activity un- derstanding. International Journal of Computer Vision , 90(1):106–129, 2010. 1

2010
[33]

Rectiﬁed linear units im- prove restricted boltzmann machines

Vinod Nair and Geoffrey E Hinton. Rectiﬁed linear units im- prove restricted boltzmann machines. In Proceedings of the 27th international conference on machine learning (ICML- 10), pages 807–814, 2010. 4

2010
[34]

Learning to rank in person re-identiﬁcation with metric ensembles

Sakrapee Paisitkriangkrai, Chunhua Shen, and Anton Van Den Hengel. Learning to rank in person re-identiﬁcation with metric ensembles. In Proceedings of the IEEE Con- ference on Computer Vision and Pattern Recognition, pages 1846–1855, 2015. 2

2015
[35]

Fast and scalable polynomial kernels via explicit feature maps

Ninh Pham and Rasmus Pagh. Fast and scalable polynomial kernels via explicit feature maps. In Proceedings of the 19th ACM SIGKDD international conference on Knowledge dis- covery and data mining, pages 239–247. ACM, 2013. 3

2013
[36]

https://pytorch.org/

Pytorch. https://pytorch.org/. 5, 6
[37]

Performance measures and a data set for multi-target, multi-camera tracking

Ergys Ristani, Francesco Solera, Roger Zou, Rita Cucchiara, and Carlo Tomasi. Performance measures and a data set for multi-target, multi-camera tracking. InEuropean Conference on Computer Vision, pages 17–35. Springer, 2016. 2, 6, 7

2016
[38]

Imagenet large scale visual recognition challenge

Olga Russakovsky, Jia Deng, Hao Su, Jonathan Krause, San- jeev Satheesh, Sean Ma, Zhiheng Huang, Andrej Karpathy, Aditya Khosla, Michael Bernstein, et al. Imagenet large scale visual recognition challenge. International journal of computer vision, 115(3):211–252, 2015. 2

2015
[39]

End-to-end deep kronecker-product matching for person re-identiﬁcation

Yantao Shen, Tong Xiao, Hongsheng Li, Shuai Yi, and Xi- aogang Wang. End-to-end deep kronecker-product matching for person re-identiﬁcation. In Proceedings of the IEEE Con- ference on Computer Vision and Pattern Recognition, pages 6886–6895, 2018. 6, 7

2018
[40]

Deep attributes driven multi-camera person re- identiﬁcation

Chi Su, Shiliang Zhang, Junliang Xing, Wen Gao, and Qi Tian. Deep attributes driven multi-camera person re- identiﬁcation. In European conference on computer vision , pages 475–491. Springer, 2016. 2

2016
[41]

Part-aligned bilinear representations for per- son re-identiﬁcation

Yumin Suh, Jingdong Wang, Siyu Tang, Tao Mei, and Ky- oung Mu Lee. Part-aligned bilinear representations for per- son re-identiﬁcation. In Proceedings of the European Con- ference on Computer Vision (ECCV), pages 402–419, 2018. 6, 7

2018
[42]

Svdnet for pedestrian retrieval

Yifan Sun, Liang Zheng, Weijian Deng, and Shengjin Wang. Svdnet for pedestrian retrieval. In Proceedings of the IEEE International Conference on Computer Vision, pages 3800– 3808, 2017. 6, 7

2017
[43]

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

Yifan Sun, Liang Zheng, Yi Yang, Qi Tian, and Shengjin Wang. Beyond part models: Person retrieval with reﬁned part pooling (and a strong convolutional baseline). In Pro- ceedings of the European Conference on Computer Vision (ECCV), pages 480–496, 2018. 2, 5, 6, 7

2018
[44]

Multiple people tracking by lifted multicut and per- son re-identiﬁcation

Siyu Tang, Mykhaylo Andriluka, Bjoern Andres, and Bernt Schiele. Multiple people tracking by lifted multicut and per- son re-identiﬁcation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , pages 3539– 3548, 2017. 1

2017
[45]

Gated siamese convolutional neural network architecture for human re-identiﬁcation

Rahul Rama Varior, Mrinal Haloi, and Gang Wang. Gated siamese convolutional neural network architecture for human re-identiﬁcation. In European Conference on Computer Vi- sion, pages 791–808. Springer, 2016. 1

2016
[46]

Attention is all you need

Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszko- reit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. In Advances in Neural Information Processing Systems, pages 5998–6008, 2017. 2

2017
[47]

Mancs: A multi-task attentional network with curriculum sampling for person re-identiﬁcation

Cheng Wang, Qian Zhang, Chang Huang, Wenyu Liu, and Xinggang Wang. Mancs: A multi-task attentional network with curriculum sampling for person re-identiﬁcation. InThe European Conference on Computer Vision (ECCV), Septem- ber 2018. 6, 7

2018
[48]

Person re-identiﬁcation by video ranking

Taiqing Wang, Shaogang Gong, Xiatian Zhu, and Shengjin Wang. Person re-identiﬁcation by video ranking. In Eu- ropean Conference on Computer Vision , pages 688–703. Springer, 2014. 2

2014
[49]

Intelligent multi-camera video surveil- lance: A review

Xiaogang Wang. Intelligent multi-camera video surveil- lance: A review. Pattern recognition letters , 34(1):3–19,
[50]

Resource aware person re-identiﬁcation across multiple resolutions

Yan Wang, Lequn Wang, Yurong You, Xu Zou, Vincent Chen, Serena Li, Gao Huang, Bharath Hariharan, and Kil- ian Q Weinberger. Resource aware person re-identiﬁcation across multiple resolutions. In Proceedings of the IEEE Con- ference on Computer Vision and Pattern Recognition, p...

2018
[51]

Person re-identiﬁcation using kernel-based metric learning methods

Fei Xiong, Mengran Gou, Octavia Camps, and Mario Sz- naier. Person re-identiﬁcation using kernel-based metric learning methods. In European conference on computer vi- sion, pages 1–16. Springer, 2014. 2

2014
[52]

Attention-aware compositional network for person re-identiﬁcation

Jing Xu, Rui Zhao, Feng Zhu, Huaming Wang, and Wanli Ouyang. Attention-aware compositional network for person re-identiﬁcation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , pages 2119– 2128, 2018. 1, 2

2018
[53]

Show, attend and tell: Neural image caption gen- eration with visual attention

Kelvin Xu, Jimmy Ba, Ryan Kiros, Kyunghyun Cho, Aaron Courville, Ruslan Salakhudinov, Rich Zemel, and Yoshua Bengio. Show, attend and tell: Neural image caption gen- eration with visual attention. In International conference on machine learning, pages 2048–2057, 2015. 2

2015
[54]

Deep metric learning for person re-identiﬁcation

Dong Yi, Zhen Lei, Shengcai Liao, and Stan Z Li. Deep metric learning for person re-identiﬁcation. In 2014 22nd International Conference on Pattern Recognition, pages 34–

2014
[55]

Harry potter’s marauder’s map: Localizing and tracking multiple persons-of-interest by nonnegative discretization

Shoou-I Yu, Yi Yang, and Alexander Hauptmann. Harry potter’s marauder’s map: Localizing and tracking multiple persons-of-interest by nonnegative discretization. In Pro- ceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 3714–3720, 2013. 1

2013
[56]

Multi- modal factorized bilinear pooling with co-attention learning for visual question answering

Zhou Yu, Jun Yu, Jianping Fan, and Dacheng Tao. Multi- modal factorized bilinear pooling with co-attention learning for visual question answering. In Proceedings of the IEEE international conference on computer vision , pages 1821– 1830, 2017. 3

2017
[57]

Beyond bilinear: generalized multimodal factorized high-order pooling for visual question answering

Zhou Yu, Jun Yu, Chenchao Xiang, Jianping Fan, and Dacheng Tao. Beyond bilinear: generalized multimodal factorized high-order pooling for visual question answering. IEEE transactions on neural networks and learning systems, (99):1–13, 2018. 3

2018
[58]

Spindle net: Person re-identiﬁcation with human body region guided feature decomposition and fusion

Haiyu Zhao, Maoqing Tian, Shuyang Sun, Jing Shao, Junjie Yan, Shuai Yi, Xiaogang Wang, and Xiaoou Tang. Spindle net: Person re-identiﬁcation with human body region guided feature decomposition and fusion. In Proceedings of the IEEE Conference on Computer Vision and Pattern Rec...

2017
[59]

Deeply-learned part-aligned representations for person re- identiﬁcation

Liming Zhao, Xi Li, Yueting Zhuang, and Jingdong Wang. Deeply-learned part-aligned representations for person re- identiﬁcation. In Proceedings of the IEEE International Conference on Computer Vision, pages 3219–3228, 2017. 2

2017
[60]

Learning mid-level ﬁlters for person re-identiﬁcation

Rui Zhao, Wanli Ouyang, and Xiaogang Wang. Learning mid-level ﬁlters for person re-identiﬁcation. In Proceed- ings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 144–151, 2014. 2

2014
[61]

Scalable person re-identiﬁcation: A benchmark

Liang Zheng, Liyue Shen, Lu Tian, Shengjin Wang, Jing- dong Wang, and Qi Tian. Scalable person re-identiﬁcation: A benchmark. In Proceedings of the IEEE International Conference on Computer Vision, pages 1116–1124, 2015. 2, 6, 7

2015
[62]

Query-adaptive late fusion for image search and person re-identiﬁcation

Liang Zheng, Shengjin Wang, Lu Tian, Fei He, Ziqiong Liu, and Qi Tian. Query-adaptive late fusion for image search and person re-identiﬁcation. In Proceedings of the IEEE con- ference on computer vision and pattern recognition , pages 1741–1750, 2015. 2

2015
[63]

Per- son re-identiﬁcation: Past, present and future

Liang Zheng, Yi Yang, and Alexander G Hauptmann. Per- son re-identiﬁcation: Past, present and future. arXiv preprint arXiv:1610.02984, 2016. 2, 5, 6, 7, 8

2016 arXiv
[64]

Re-identiﬁcation with consistent attentive siamese networks

Meng Zheng, Srikrishna Karanam, Ziyan Wu, and Richard J Radke. Re-identiﬁcation with consistent attentive siamese networks. In Proceedings of the IEEE conference on com- puter vision and pattern recognition, 2019. 6, 7, 8

2019
[65]

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

Zhedong Zheng, Liang Zheng, and Yi Yang. Unlabeled sam- ples generated by gan improve the person re-identiﬁcation baseline in vitro. In Proceedings of the IEEE International Conference on Computer Vision, pages 3754–3762, 2017. 2, 6, 7

2017
[66]

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

Zhun Zhong, Liang Zheng, Donglin Cao, and Shaozi Li. Re- ranking person re-identiﬁcation with k-reciprocal encoding. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 1318–1327, 2017. 2, 6, 7

2017
[67]

Random erasing data augmentation

Zhun Zhong, Liang Zheng, Guoliang Kang, Shaozi Li, and Yi Yang. Random erasing data augmentation. arXiv preprint arXiv:1708.04896, 2017. 6

2017 arXiv

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

https://github.com/layumi/Person_reID_ baseline_pytorch. 6

work page

[2] [2]

Higher-order in- tegration of hierarchical convolutional activations for ﬁne- grained visual categorization

Sijia Cai, Wangmeng Zuo, and Lei Zhang. Higher-order in- tegration of hierarchical convolutional activations for ﬁne- grained visual categorization. In Proceedings of the IEEE International Conference on Computer Vision , pages 511– 520, 2017. 3

work page 2017

[3] [3]

Multi-level factorisation net for person re-identiﬁcation

Xiaobin Chang, Timothy M Hospedales, and Tao Xiang. Multi-level factorisation net for person re-identiﬁcation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 2109–2118, 2018. 6, 7

work page 2018

[4] [4]

Synthesized classiﬁers for zero-shot learning

Soravit Changpinyo, Wei-Lun Chao, Boqing Gong, and Fei Sha. Synthesized classiﬁers for zero-shot learning. In Pro- ceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 5327–5336, 2016. 3

work page 2016

[5] [5]

Energy confused adver- sarial metric learning for zero-shot image retrieval and clus- tering

Binghui Chen and Weihong Deng. Energy confused adver- sarial metric learning for zero-shot image retrieval and clus- tering. In AAAI Conference on Artiﬁcial Intelligence, 2019. 2, 3, 5

work page 2019

[6] [6]

Hybrid-attention based decoupled metric learning for zero-shot image retrieval

Binghui Chen and Weihong Deng. Hybrid-attention based decoupled metric learning for zero-shot image retrieval. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 2750–2759, 2019. 3

work page 2019

[7] [7]

Noisy soft- max: Improving the generalization ability of dcnn via post- poning the early softmax saturation

Binghui Chen, Weihong Deng, and Junping Du. Noisy soft- max: Improving the generalization ability of dcnn via post- poning the early softmax saturation. In The IEEE Confer- ence on Computer Vision and Pattern Recognition (CVPR) , July 2017. 2

work page 2017

[8] [8]

Vir- tual class enhanced discriminative embedding learning

Binghui Chen, Weihong Deng, and Haifeng Shen. Vir- tual class enhanced discriminative embedding learning. In Advances in Neural Information Processing Systems , pages 1946–1956, 2018. 2

work page 1946