Pith. sign in

REVIEW 3 major objections 4 minor 50 references

An End-to-End Framework for Video Multi-Person Pose Estimation

T0 review · 3 major / 4 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read VEPE treats a video clip as a sequence-to-sequence pose problem: on PoseTrack2017 it reaches 83.0 mAP at 334 ms per frame, beating DCPose's 82.8 mAP at 1390 ms per frame.

desk verdict A coherent PETR extension with honest ablations, but the manuscript is incomplete: two comparison tables are missing and the 300% speedup claim rests on uncontrolled cross-paper timings. read the letter →

arxiv 2509.01095 v1 pith:NZMJBYVU submitted 2025-09-01 cs.CV

Zhihong Wei This is my paper
classification cs.CV
keywords videomulti-personposeestimationend-to-endTransformerspatio-temporalattentiondeformableinstanceconsistencyTrackquerytemporalmodeling
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

VEPE is an attempt to make video multi-person pose estimation fully end-to-end: instead of detecting and cropping each person per frame and then running a single-person temporal model, it treats the whole clip as a sequence-to-sequence problem solved by Transformer queries. The paper's central claim is that three spatio-temporal modules—a pose encoder, a deformable memory encoder, and a cascaded pose decoder—plus an instance-consistency loss can aggregate temporal context well enough to match or beat two-stage pipelines while removing detectors, RoI cropping, and NMS. On PoseTrack2017, VEPE reports 83.0 mAP at 334 ms per frame, compared with DCPose's 82.8 mAP at 1390 ms, which the paper summarizes as about a 300% inference-efficiency gain. If true, accurate video pose estimation does not need the expensive two-stage apparatus, and inference time is tied to a fixed set of pose queries rather than to the number of people in the frame.

What carries the argument

Four objects carry the argument. (1) STPE: self-attention over keyframe pose queries plus cross-attention with reference-frame queries, masked by instance-query similarity so a query attends to its match. (2) STDME: temporal multi-scale deformable attention that samples T×L×K points across frames and scales, pushing cross-frame visual cues into the keyframe feature memory. (3) STPD: three cascaded decoder layers refining normalized coordinates by Q_d = σ(σ^{-1}(Q_{d-1}) + ΔQ_p). (4) Instance consistency mechanism: learnable instance queries, one per pose query, trained with L_ic = Σ_i max(0, d(a_i,p_i) − d(a_i,n_i) + margin); the same similarity scores give tracking and the STPE attention ma

What would settle it

Run VEPE and DCPose on the same machine, same software stack, same input resolution (shortest edge 800), and same batch size, timing PoseTrack2017 per frame on a single V100; if VEPE does not reproduce roughly 83.0 mAP at roughly 334 ms while DCPose sits near 82.8 mAP at 1390 ms, the central claim fails.

Watch

Extended reading notes

Core claim

On its own terms, the discovery is that temporal context for multi-person pose estimation can be folded into a single query-based Transformer. VEPE extends PETR with three spatio-temporal modules that align same-instance pose queries, aggregate cross-frame multi-scale visual features, and refine keyframe poses. A margin-based instance-consistency loss keeps the same person's instance queries close across frames and different people's apart; similarity between instance queries supplies tracking and an attention mask. The reported result: 83.0 mAP on PoseTrack2017 validation, up 5.8 over the spatial baseline, slightly above DCPose, at 334 ms per frame vs 1390 ms for DCPose.

Load-bearing premise

The 300% speed gain assumes that per-frame inference times reported by different papers are directly comparable because they were all tested on a V100 GPU, without controlling for software versions, batching, input resolution, or model parameter counts.

Editorial extensions

If this is right

  • Video pose estimation can drop explicit person detection, RoI cropping, and NMS without sacrificing accuracy to two-stage models, at least on PoseTrack-style benchmarks.
  • The temporal modules are worth about 5.8 mAP over the spatial-only PETR baseline on PoseTrack2017, with STPE contributing the largest single gain in the ablation.
  • Because decoding cost is tied to a fixed number of pose queries rather than the number of detected persons, per-frame latency should grow much more slowly in crowded scenes than top-down pipelines' latency.
  • Instance queries learned for matching supply tracking as a by-product: similarity between instance queries identifies the same person across frames and guides cross-frame pose-query matching.
  • If the 334 ms/frame figure reproduces under controlled conditions, an end-to-end video pose estimator can operate in the same speed class as image pose models while still exploiting temporal context.

Reading between the lines

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

  • The instance-consistency loss is a generic contrastive mechanism—learned per-identity queries plus a margin-based pull/push loss—so it could be lifted to other query-based video tasks such as multi-object tracking or video instance segmentation.
  • VEPE is trained and evaluated with a fixed window of 3 frames, so it remains untested whether the STPE/STDME modules keep helping over longer clips without retraining; extending the window is a natural stress test.
  • The paper's own spatial baseline (77.2 mAP) is the evidence for the temporal modules' gain; comparing against a stronger single-frame end-to-end model would further isolate what the spatio-temporal attention adds.
  • If the efficiency advantage holds in dense scenes, end-to-end query decoders could become the default choice for real-time multi-person video pose estimation, because one network handles spatial and temporal context simultaneously.
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.

Referee Report

3 major / 4 minor

Summary. The paper proposes VEPE, an end-to-end Transformer framework for video multi-person pose estimation, extending PETR with three spatio-temporal modules (STPE, STDME, STPD) and an instance consistency mechanism (ICM) with a learnable instance query. The method is evaluated on PoseTrack2017, and the authors claim it outperforms most two-stage models on PoseTrack datasets and improves inference efficiency by 300%. The manuscript includes PoseTrack2017 validation results (83.0 mAP, 334 ms/frame), component ablations, and a threshold study for pose query selection.

Significance. If the claims are validated, VEPE would be a meaningful step toward efficient end-to-end video pose estimation: it removes the separate detector and post-processing (NMS, RoI cropping) used by two-stage methods, and the proposed temporal modules are architecturally reasonable. The PoseTrack2017 validation result is competitive with DCPose, and the ablation study suggests each component contributes. However, the current evidence is incomplete: two of the three dataset comparisons are missing, and the headline 300% efficiency claim rests on uncontrolled cross-paper timings. These issues are fixable and do not invalidate the architectural idea, but they must be addressed before the central claims can be accepted.

major comments (3)
  1. [§4.3.2–4.3.3] The text refers to 'table ??' for the PoseTrack2018 and PoseTrack2021 comparisons, but no such tables are present in the manuscript. Since the abstract and conclusion claim superiority 'on the Posetrack dataset' and the conclusion repeats the claim, these results are load-bearing. Without the actual PoseTrack2018 and PoseTrack2021 numbers, the accuracy component of the central claim is unverifiable on two of the three datasets. Please include the tables or restrict the claims explicitly to PoseTrack2017.
  2. [§4.2.3, Table 1] The 300% inference-efficiency claim is based on comparing VEPE's 334 ms/frame with DCPose's 1390 ms/frame and FAMI-Pose's 1577 ms/frame. The text only states that all methods were tested on a V100 GPU. This does not control for software versions, batching, input resolution, whether timings for top-down baselines include detector and post-processing, parameter counts, or measurement protocols. As reported, the speed advantage is not established. Provide a controlled benchmark on identical hardware/software, or rephrase the claim as a reported-per-paper comparison with explicit caveats.
  3. [Table 1, §4.2.3] All baselines in Table 1 except VEPE are two-stage models. Since the paper's stated contribution is an end-to-end framework and the efficiency argument is framed against two-stage pipelines, the comparison should also include recent end-to-end pose estimators (PETR, QueryPose, ED-POSE, GroupPose) on the same benchmark. Without such comparisons, the efficiency and accuracy advantages relative to the end-to-end family are not demonstrated.
minor comments (4)
  1. [§4.4.1–4.4.2] The relationship between Pose Query Selection (PQS) and the STPE ablation is unclear. Table 2 reports that adding STPE raises mAP from 77.2 to 80.7, and Table 3 reports threshold effects around 80.4–80.7. Is PQS included in the STPE row? What threshold is used for the full model in Table 1? Please clarify and consider adding a PQS row/column to the ablation table.
  2. [§3.6, Eq. (5)] The instance consistency loss uses a triplet formulation with cosine distance, but the construction of the candidate set, the choice of negatives, and the handling of multiple negatives are not specified. Please define how anchors, positives, and negatives are sampled across frames and how the loss is aggregated over the video clip.
  3. [§4.4.3, Figure 6] The instance tracking function is only supported by a qualitative similarity heatmap. No quantitative tracking metric (e.g., MOTA, IDF1) is reported. Either report such a metric or restrict the tracking claim to its effect on pose estimation mAP.
  4. [Throughout] There are several typos and inconsistent notations: 'HMA' in Eq. (2) should be 'MHA'; 'memroy' after Eq. (3); 'Posetrack' vs 'PoseTrack'; 'the row column' in the Figure 5 caption. Please proofread carefully.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: VEPE's claims are empirical benchmark results; the instance-consistency loss is a standard supervised metric-learning objective.

full rationale

VEPE's central claims are empirical: an end-to-end video pose framework whose components (STPE, STDME, STPD, ICM) are described architecturally and evaluated on PoseTrack2017/2018/2021. The architecture builds on PETR and deformable attention, but it does not derive a prediction from a fitted parameter that is itself the target. The one mechanism that might look self-referential is the Instance Consistency Mechanism: Eq. (5) supervises cosine similarity between instance queries using ground-truth trackId labels to pull same-person queries together and push different-person queries apart, and the resulting similarity is then used to build an instance mask for cross-frame matching. This is a standard metric-learning / triplet-style training loop, not a case where the reported outcome is equivalent to the training signal by construction; the evaluation metric is pose mAP, not the learned similarity itself. The headline efficiency gain (334 ms vs. 1390 ms) rests on cross-paper timing comparisons with only a V100 GPU stated in Section 4.2.3, which is a measurement-protocol risk rather than circularity. Also, the missing 'Table ??' references for PoseTrack2018 and PoseTrack2021 are verifiability/completeness defects, not circular derivations. There are no load-bearing self-citations, no imported uniqueness theorems, and no renamed known result presented as a derivation. Accordingly, no specific circular step can be quoted and exhibited, so the correct finding is no significant circularity.

Assumptions & free parameters 4 free parameters · 3 assumptions · 1 invented entities

The paper's central claims depend on the quality of PoseTrack annotations, the correctness of base implementations (PETR, HRNet, Deformable DETR), and several hand-set hyperparameters. No code is released, so these assumptions cannot be independently checked.

free parameters (4)
  • Pose Query Selection confidence threshold = 0.3
    Chosen based on validation sweep in Table 3; mAP varies from 80.1 to 80.7 across thresholds 0.1-0.5.
  • Instance consistency loss margin = not stated
    Eq. 5 includes a margin hyperparameter that controls positive/negative separation; the paper never reports its value.
  • Number of pose queries = 100
    Initialized to 100, following PETR; affects model capacity and matching cost.
  • Number of temporal frames = 3
    Fixed to 3 frames to match prior models; not swept or justified beyond comparability.
assumptions (3)
  • domain assumption PoseTrack annotations are accurate and used as ground truth
    All mAP numbers depend on the quality of PoseTrack labels; no annotation quality analysis is provided.
  • domain assumption PETR, Deformable DETR, and HRNet behave as described in their original papers
    VEPE builds directly on these components; if the base implementations differ, results change.
  • domain assumption Cosine similarity between learned instance query embeddings is a valid proxy for same-person identity across frames
    The instance mask and loss assume similar embeddings correspond to the same person; this is a modeling assumption not proven.
invented entities (1)
  • Instance query
    purpose: A learnable embedding per pose query used to associate the same person across frames and to filter attention via an instance mask.
    It is an internal representation with no observable prediction outside the model; its effectiveness is only validated through end-task mAP.

how reviews work

0 comments
Cite this review

Pith. "Pith review of An End-to-End Framework for Video Multi-Person Pose Estimation." pith.science (2026). https://pith.science/paper/NZMJBYVU

@misc{pith2026250901095,
  author       = {Pith},
  title        = {Pith review of: An End-to-End Framework for Video Multi-Person Pose Estimation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NZMJBYVU}},
  note         = {Machine review of arXiv:2509.01095}
}
read the original abstract

Video-based human pose estimation models aim to address scenarios that cannot be effectively solved by static image models such as motion blur, out-of-focus and occlusion. Most existing approaches consist of two stages: detecting human instances in each image frame and then using a temporal model for single-person pose estimation. This approach separates the spatial and temporal dimensions and cannot capture the global spatio-temporal context between spatial instances for end-to-end optimization. In addition, it relies on separate detectors and complex post-processing such as RoI cropping and NMS, which reduces the inference efficiency of the video scene. To address the above problems, we propose VEPE (Video End-to-End Pose Estimation), a simple and flexible framework for end-to-end pose estimation in video. The framework utilizes three crucial spatio-temporal Transformer components: the Spatio-Temporal Pose Encoder (STPE), the Spatio-Temporal Deformable Memory Encoder (STDME), and the Spatio-Temporal Pose Decoder (STPD). These components are designed to effectively utilize temporal context for optimizing human body pose estimation. Furthermore, to reduce the mismatch problem during the cross-frame pose query matching process, we propose an instance consistency mechanism, which aims to enhance the consistency and discrepancy of the cross-frame instance query and realize the instance tracking function, which in turn accurately guides the pose query to perform cross-frame matching. Extensive experiments on the Posetrack dataset show that our approach outperforms most two-stage models and improves inference efficiency by 300%.

Figures

Figures reproduced from arXiv: 2509.01095 by the authors.

Figure 1
Figure 1. Comparison of multi-stage and end-to-end frameworks. [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. VEPE model pipeline. The VEPE model pipeline consists of two phases: spatial and temporal. In the spatial phase: spatial [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. The structure of (a). Spatio-Temporal Pose Encoder [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Detailed Architecture of STPD. STPD employs a self [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Visualization of Posetrack21 validation results. The first [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Instance tracking between cross-frame instance queries [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

50 extracted references · 45 canonical work pages

  1. [1]

    Posetrack: A benchmark for human pose estima- tion and tracking

    Mykhaylo Andriluka, Umar Iqbal, Eldar Insafutdinov, Leonid Pishchulin, Anton Milan, Juergen Gall, and Bernt Schiele. Posetrack: A benchmark for human pose estima- tion and tracking. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 5167–5176,

    work page

  2. [2]

    Unipose: Unified hu- man pose estimation in single images and videos

    Bruno Artacho and Andreas Savakis. Unipose: Unified hu- man pose estimation in single images and videos. In Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 7035–7044, 2020. 2, 3

    work page 2020

  3. [3]

    Pose-guided tracking-by-detection: Robust multi- person pose tracking

    Qian Bao, Wu Liu, Yuhao Cheng, Boyan Zhou, and Tao Mei. Pose-guided tracking-by-detection: Robust multi- person pose tracking. IEEE Transactions on Multimedia, 23: 161–175, 2020. 8

    work page 2020

  4. [4]

    Tracking without bells and whistles

    Philipp Bergmann, Tim Meinhardt, and Laura Leal-Taixe. Tracking without bells and whistles. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pages 941–951, 2019. 8

    work page 2019

  5. [5]

    Learning temporal pose esti- mation from sparsely-labeled videos

    Gedas Bertasius, Christoph Feichtenhofer, Du Tran, Jianbo Shi, and Lorenzo Torresani. Learning temporal pose esti- mation from sparsely-labeled videos. In Advances in Neural Information Processing Systems, pages 3027–3038, 2019. 3, 8

    work page 2019

  6. [6]

    End-to- end object detection with transformers

    Nicolas Carion, Francisco Massa, Gabriel Synnaeve, Nicolas Usunier, Alexander Kirillov, and Sergey Zagoruyko. End-to- end object detection with transformers. In European Confer- ence on Computer Vision , pages 213–229. Springer, 2020. 3

    work page 2020

  7. [7]

    Multi-context attention for hu- man pose estimation

    Xiao Chu, Wei Yang, Wanli Ouyang, Cheng Ma, Alan L Yuille, and Xiaogang Wang. Multi-context attention for hu- man pose estimation. InProceedings of the IEEE Conference on Computer Vision and Pattern Recognition , pages 1831– 1840, 2017. 1

    work page 2017

  8. [8]

    Joint flow: Temporal flow fields for multi person tracking

    Andreas Doering, Umar Iqbal, and Juergen Gall. Joint flow: Temporal flow fields for multi person tracking. arXiv preprint arXiv:1805.04596, 2018. 8

    arXiv 2018

Show all 50 references
  1. [9]

    Posetrack21: A dataset for person search, multi-object tracking and multi-person pose tracking

    Andreas Doering, Di Chen, Shanshan Zhang, Bernt Schiele, and Juergen Gall. Posetrack21: A dataset for person search, multi-object tracking and multi-person pose tracking. InPro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 20963–20972, ...

    2022

  2. [10]

    Rmpe: Regional multi-person pose estimation

    Hao-Shu Fang, Shuqin Xie, Yu-Wing Tai, and Cewu Lu. Rmpe: Regional multi-person pose estimation. In Proceed- ings of the IEEE International Conference on Computer Vi- sion, pages 2334–2343, 2017. 1, 8

    2017

  3. [11]

    Detect-and-track: Efficient pose estimation in videos

    Rohit Girdhar, Georgia Gkioxari, Lorenzo Torresani, Manohar Paluri, and Du Tran. Detect-and-track: Efficient pose estimation in videos. In Proceedings of the IEEE Con- ference on Computer Vision and Pattern Recognition, pages 350–359, 2018. 8

    2018

  4. [12]

    Multi-domain pose net- work for multi-person pose estimation and tracking

    Hengkai Guo, Tang Tang, Guozhong Luo, Riwei Chen, Yongchen Lu, and Linfu Wen. Multi-domain pose net- work for multi-person pose estimation and tracking. In Pro- ceedings of the European Conference on Computer Vision (ECCV), pages 0–0, 2018. 8

    2018

  5. [13]

    Pose estimator and tracker using temporal flow maps for limbs

    Jihye Hwang, Jieun Lee, Sungheon Park, and Nojun Kwak. Pose estimator and tracker using temporal flow maps for limbs. In 2019 International Joint Conference on Neural Networks (IJCNN), pages 1–8. IEEE, 2019. 8

    2019

  6. [14]

    Posetrack: Joint multi-person pose estimation and tracking

    Umar Iqbal, Anton Milan, and Juergen Gall. Posetrack: Joint multi-person pose estimation and tracking. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 2011–2020, 2017. 7

    2011

  7. [15]

    Rtmpose: Real- time multi-person pose estimation based on mmpose

    Tao Jiang, Peng Lu, Li Zhang, Ningsheng Ma, Rui Han, Chengqi Lyu, Yining Li, and Kai Chen. Rtmpose: Real- time multi-person pose estimation based on mmpose. arXiv preprint arXiv:2303.07399, 2023. 1

    2023 arXiv

  8. [16]

    Multi-person articulated tracking with spatial and tempo- ral embeddings

    Sheng Jin, Wentao Liu, Wanli Ouyang, and Chen Qian. Multi-person articulated tracking with spatial and tempo- ral embeddings. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , pages 5664– 5673, 2019. 8

    2019

  9. [17]

    The hungarian method for the assignment problem

    Harold W Kuhn. The hungarian method for the assignment problem. Naval research logistics quarterly , 2(1-2):83–97,

  10. [18]

    Simcc: A simple coordinate classification perspective for hu- man pose estimation

    Yanjie Li, Sen Yang, Peidong Liu, Shoukui Zhang, Yunx- iao Wang, Zhicheng Wang, Wankou Yang, and Shu-Tao Xia. Simcc: A simple coordinate classification perspective for hu- man pose estimation. In European Conference on Computer Vision, pages 89–106. Springer, 2022. 1

    2022

  11. [19]

    Group pose: A simple baseline for end-to- end multi-person pose estimation

    Huan Liu, Qiang Chen, Zichang Tan, Jiang-Jiang Liu, Jian Wang, Xiangbo Su, Xiaolong Li, Kun Yao, Junyu Han, Er- rui Ding, et al. Group pose: A simple baseline for end-to- end multi-person pose estimation. In Proceedings of the IEEE/CVF International Conference on Computer Visi...

    2023

  12. [20]

    Towards natural and accurate future motion prediction of humans and animals

    Zhenguang Liu, Shuang Wu, Shuyuan Jin, Qi Liu, Shijian Lu, Roger Zimmermann, and Li Cheng. Towards natural and accurate future motion prediction of humans and animals. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 10004–10012, 2019. 1

    2019

  13. [21]

    Deep dual consec- utive network for human pose estimation

    Zhenguang Liu, Haoming Chen, Runyang Feng, Shuang Wu, Shouling Ji, Bailin Yang, and Xun Wang. Deep dual consec- utive network for human pose estimation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 525–534, 2021. 2, 3, 8

    2021

  14. [22]

    Tempo- ral feature alignment and mutual information maximization for video-based human pose estimation

    Zhenguang Liu, Runyang Feng, Haoming Chen, Shuang Wu, Yixing Gao, Yunjun Gao, and Xiang Wang. Tempo- ral feature alignment and mutual information maximization for video-based human pose estimation. In Proceedings of the IEEE/CVF conference on computer vision and pattern recogn...

    2022

  15. [23]

    Lstm pose ma- chines

    Yue Luo, Jimmy Ren, Zhouxia Wang, Wenxiu Sun, Jinshan Pan, Jianbo Liu, Jiahao Pang, and Liang Lin. Lstm pose ma- chines. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 5207–5215, 2018. 1, 2, 3

    2018

  16. [24]

    Stacked hour- glass networks for human pose estimation

    Alejandro Newell, Kaiyu Yang, and Jia Deng. Stacked hour- glass networks for human pose estimation. InEuropean con- ference on computer vision, pages 483–499. Springer, 2016. 1

    2016

  17. [25]

    Flow- ing convnets for human pose estimation in videos

    Tomas Pfister, James Charles, and Andrew Zisserman. Flow- ing convnets for human pose estimation in videos. In Pro- ceedings of the IEEE International Conference on Computer Vision, pages 1913–1921, 2015. 2, 3

    1913

  18. [26]

    Efficient online multi-person 2d pose tracking with recurrent spatio-temporal affinity fields

    Yaadhav Raaj, Haroon Idrees, Gines Hidalgo, and Yaser Sheikh. Efficient online multi-person 2d pose tracking with recurrent spatio-temporal affinity fields. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 4620–4628, 2019. 8

    2019

  19. [27]

    Self-supervised keypoint correspondences for multi- person pose estimation and tracking in videos

    Umer Rafi, Andreas Doering, Bastian Leibe, and Juergen Gall. Self-supervised keypoint correspondences for multi- person pose estimation and tracking in videos. In European Conference on Computer Vision , pages 36–52. Springer,

  20. [28]

    Yolov3: An incremental improvement

    Joseph Redmon and Ali Farhadi. Yolov3: An incremental improvement. arXiv preprint arXiv:1804.02767, 2018. 7, 8

    2018 arXiv

  21. [29]

    End-to-end multi-person pose estimation with transformers

    Dahu Shi, Xing Wei, Liangqi Li, Ye Ren, and Wenming Tan. End-to-end multi-person pose estimation with transformers. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 11069–11078, 2022. 2, 3, 7

    2022

  22. [30]

    Thin-slicing network: A deep structured model for pose esti- mation in videos

    Jie Song, Limin Wang, Luc Van Gool, and Otmar Hilliges. Thin-slicing network: A deep structured model for pose esti- mation in videos. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 4220–4229,

  23. [31]

    Deep high-resolution representation learning for human pose esti- mation

    Ke Sun, Bin Xiao, Dong Liu, and Jingdong Wang. Deep high-resolution representation learning for human pose esti- mation. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 5693–5703, 2019. 7, 8

    2019

  24. [32]

    Sparse r-cnn: End-to-end ob- ject detection with learnable proposals

    Peize Sun, Rufeng Zhang, Yi Jiang, Tao Kong, Chen- feng Xu, Wei Zhan, Masayoshi Tomizuka, Lei Li, Zehuan Yuan, Changhu Wang, et al. Sparse r-cnn: End-to-end ob- ject detection with learnable proposals. In Proceedings of the IEEE/CVF conference on computer vision and pattern re...

    2021

  25. [33]

    Deeppose: Human pose estimation via deep neural networks

    Alexander Toshev and Christian Szegedy. Deeppose: Human pose estimation via deep neural networks. In Proceedings of the IEEE conference on computer vision and pattern recog- nition, pages 1653–1660, 2014. 1

    2014

  26. [34]

    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. 5

    2017

  27. [35]

    Beyond physical connections: Tree models in human pose estimation

    Fang Wang and Yi Li. Beyond physical connections: Tree models in human pose estimation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recog- nition, pages 596–603, 2013. 1

    2013

  28. [36]

    Deep high-resolution represen- tation learning for visual recognition

    Jingdong Wang, Ke Sun, Tianheng Cheng, Borui Jiang, Chaorui Deng, Yang Zhao, Dong Liu, Yadong Mu, Mingkui Tan, Xinggang Wang, et al. Deep high-resolution represen- tation learning for visual recognition. IEEE transactions on pattern analysis and machine intelligence, 2020. 1

    2020

  29. [37]

    Com- bining detection and tracking for human pose estimation in videos

    Manchen Wang, Joseph Tighe, and Davide Modolo. Com- bining detection and tracking for human pose estimation in videos. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 11088– 11096, 2020. 1, 2, 3

    2020

  30. [38]

    Convolutional pose machines

    Shih-En Wei, Varun Ramakrishna, Takeo Kanade, and Yaser Sheikh. Convolutional pose machines. In Proceedings of the IEEE conference on Computer Vision and Pattern Recogni- tion, pages 4724–4732, 2016. 1

    2016

  31. [39]

    Simple baselines for human pose estimation and tracking

    Bin Xiao, Haiping Wu, and Yichen Wei. Simple baselines for human pose estimation and tracking. In Proceedings of the European Conference on Computer Vision (ECCV), 2018. 1

    2018

  32. [40]

    Simple baselines for human pose estimation and tracking

    Bin Xiao, Haiping Wu, and Yichen Wei. Simple baselines for human pose estimation and tracking. In Proceedings of the European conference on computer vision (ECCV), pages 466–481, 2018. 8

    2018

  33. [41]

    Querypose: Sparse multi- person pose regression via spatial-aware part-level query

    Yabo Xiao, Kai Su, Xiaojuan Wang, Dongdong Yu, Lei Jin, Mingshu He, and Zehuan Yuan. Querypose: Sparse multi- person pose regression via spatial-aware part-level query. Advances in Neural Information Processing Systems , 35: 12464–12477, 2022. 3

    2022

  34. [42]

    Pose flow: Efficient online pose tracking

    Yuliang Xiu, Jiefeng Li, Haoyu Wang, Yinghong Fang, and Cewu Lu. Pose flow: Efficient online pose tracking. arXiv preprint arXiv:1802.00977, 2018. 8

    2018 arXiv

  35. [43]

    Vit- pose: Simple vision transformer baselines for human pose estimation

    Yufei Xu, Jing Zhang, Qiming Zhang, and Dacheng Tao. Vit- pose: Simple vision transformer baselines for human pose estimation. Advances in Neural Information Processing Sys- tems, 35:38571–38584, 2022. 1

    2022

  36. [44]

    Spatial tempo- ral graph convolutional networks for skeleton-based action recognition

    Sijie Yan, Yuanjun Xiong, and Dahua Lin. Spatial tempo- ral graph convolutional networks for skeleton-based action recognition. In Proceedings of the AAAI conference on arti- ficial intelligence, 2018. 1

    2018

  37. [45]

    Explicit box detection unifies end-to-end multi-person pose estimation

    Jie Yang, Ailing Zeng, Shilong Liu, Feng Li, Ruimao Zhang, and Lei Zhang. Explicit box detection unifies end-to-end multi-person pose estimation. arXiv preprint arXiv:2302.01593, 2023. 3

    2023 arXiv

  38. [46]

    Trans- pose: Keypoint localization via transformer

    Sen Yang, Zhibin Quan, Mu Nie, and Wankou Yang. Trans- pose: Keypoint localization via transformer. In Proceedings of the IEEE/CVF International Conference on Computer Vi- sion, pages 11802–11812, 2021. 1

    2021

  39. [47]

    Learning dynamics via graph neural networks for human pose estimation and tracking

    Yiding Yang, Zhou Ren, Haoxiang Li, Chunluan Zhou, Xin- chao Wang, and Gang Hua. Learning dynamics via graph neural networks for human pose estimation and tracking. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 8074–8084, 2021. 8

    2021

  40. [48]

    Fastpose: Towards real-time pose estimation and tracking via scale-normalized multi-task net- works

    Jiabin Zhang, Zheng Zhu, Wei Zou, Peng Li, Yanwei Li, Hu Su, and Guan Huang. Fastpose: Towards real-time pose estimation and tracking via scale-normalized multi-task net- works. arXiv preprint arXiv:1908.05593, 2019. 8

    1908 arXiv

  41. [49]

    Efficient human pose estimation via parsing a tree structure based human model

    Xiaoqin Zhang, Changcheng Li, Xiaofeng Tong, Weiming Hu, Steve Maybank, and Yimin Zhang. Efficient human pose estimation via parsing a tree structure based human model. In 2009 IEEE 12th International Conference on Computer Vision, pages 1349–1356. IEEE, 2009. 1

    2009

  42. [50]

    Deformable detr: Deformable trans- formers for end-to-end object detection

    Xizhou Zhu, Weijie Su, Lewei Lu, Bin Li, Xiaogang Wang, and Jifeng Dai. Deformable detr: Deformable trans- formers for end-to-end object detection. arXiv preprint arXiv:2010.04159, 2020. 3, 5

    2010 arXiv

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.