Pith. sign in

REVIEW 4 major objections 6 minor 43 references

Detection, Pose Estimation and Segmentation for Multiple Bodies: Closing the Virtuous Circle

T0 review · 4 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read BBox-Mask-Pose closes a feedback loop among detectors, masks, and pose estimators to push multi-person perception to new state-of-the-art results in crowded scenes.

desk verdict MaskPose and the BMP loop are a genuine new contribution with strong OCHuman gains, but the pose-SOTA claim and mask-quality sensitivity need tightening before this is fully convincing. read the letter →

arxiv 2412.01562 v3 pith:LFZLH6PT submitted 2024-12-02 cs.CV

Miroslav Purkrabek , Jiri Matas This is my paper · ORCID
classification cs.CV
keywords multi-personposeestimationinstancesegmentationobjectdetectioncrowdedscenesBBox-Mask-Posemask-conditionedSAMpromptingmutualconditioning
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

Human pose estimators that work well on isolated people often fail when bodies overlap, because a bounding-box crop can contain several people and a detector can merge them into one. This paper argues that the remedy is to stop treating boxes, masks, and poses as a one-way pipeline and instead let them correct each other in a loop. The proposed BBox-Mask-Pose (BMP) method runs a detector, a mask-conditioned pose estimator, and a pose-prompted segmenter repeatedly, masking out each processed person and re-detecting the rest. On OCHuman, a benchmark dominated by heavily overlapping people, BMP claims new state-of-the-art results in detection, instance segmentation, and pose estimation without training on OCHuman, and it matches the best top-down pose results on COCO. If the paper is right, small specialized models connected by explicit mutual conditioning can beat much larger single models on crowded scenes.

What carries the argument

The load-bearing mechanism is the closed loop of Eqs. (7)-(9): a detector fine-tuned to ignore previously masked instances, MaskPose conditioned by the masked image and bounding box (with mixing parameter $\alpha = 0.8$), and SAM2 prompted by keypoints selected by a confidence-and-spread algorithm. A pose-mask consistency score (Eq. 6) discards SAM refinements whose masks are less consistent with the keypoints than the detector's original mask. The loop repeats until no new detections appear; two iterations are used in the reported results.

What would settle it

Run the two-iteration BMP loop on a held-out set of heavily overlapping people with identical clothing, and compare it with the single-pass detector and with a version in which SAM masks are replaced by ground-truth masks; if the gains persist only with ground-truth masks, the mask-out step is the bottleneck.

Watch

Extended reading notes

Core claim

The central claim is that mutual conditioning of bounding boxes, instance masks, and keypoint poses produces better estimates of all three than any single model produces alone. The paper introduces MaskPose, a top-down model that takes a semi-transparent masked image as additional conditioning; trained on COCO, MPII, and AIC, it raises OCHuman pose AP from ViTPose's 42.6 to 45.0 while keeping COCO val AP at 76.5. Feeding MaskPose's keypoints to SAM2 as automatically selected prompts, then masking out the refined instances and rerunning the detector, yields 49.2 pose AP, 35.9 bbox AP, and 34.0 mask AP on OCHuman after two iterations, beating the prior OCHuman pose SOTA of 48.3. The same loop matches top-down pose SOTA on COCO and improves detection most in scenes with large bounding-box overlap. The paper also reports that the loop's effect saturates after two iterations and that most remaining errors come from imperfect SAM masks.

Load-bearing premise

The loop assumes that each refined instance mask covers exactly one visible person, so blacking it out neither hides a second person nor leaves clothing and limbs behind to be re-detected as new instances.

Editorial extensions

If this is right

  • Mask conditioning of top-down pose models improves crowded-scene pose without hurting standard benchmarks: MaskPose raises OCHuman AP from 42.6 to 45.0 while holding COCO val AP at 76.5.
  • Iterating the loop recovers instances a one-pass detector misses; detection AP on OCHuman rises from 30.0 for RTMDet-L to 35.9 after two BMP iterations.
  • The gains concentrate where bounding boxes overlap heavily; BMP improves detection by 5.8 AP for instances with max IoU between 0.6 and 0.8 and by about 39% overall in large-overlap scenes.
  • Two iterations are sufficient; further iterations add computational cost without meaningful gains.
  • Three small specialized models (about 369M parameters, 1.12 s/image for two iterations) can outperform a much larger human foundation model (about 336M parameters alone, 1.95 s/image with a detector) in both accuracy and speed.

Reading between the lines

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

  • A natural extension the paper does not test is applying the same mutual-consistency loop to other articulated object classes, such as animals or vehicles, that have detector, segmenter, and keypoint models.
  • Because bounding-box prompting helps once boxes are stable, a post-loop refinement pass that prompts SAM with the final bounding boxes could further improve masks; the paper mentions this option but leaves it out of its reported numbers.
  • The paper's finding that lower confidence thresholds help SAM prompting suggests that calibrated keypoint uncertainty, such as a predicted OKS, could make automated prompting more reliable than raw confidence.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper proposes BBox-Mask-Pose (BMP), an iterative loop that couples a fine-tuned RTMDet detector, a new mask-conditioned top-down pose estimator (MaskPose), and SAM2. The loop alternately detects persons, estimates poses, refines instance masks, masks out processed instances, and re-detects. The authors report improved detection, pose, and segmentation on OCHuman (e.g., 35.9 bbox AP, 34.0 mask AP, 49.2 pose AP at two iterations) and claim state-of-the-art pose accuracy on COCO (76.5 val AP), without training any component on OCHuman and with the same hyperparameters on both datasets. The core idea is that enforcing mutual consistency among bounding boxes, masks, and poses closes a 'virtuous circle' that helps especially when overlapping bodies are merged by a detector.

Significance. If the reported numbers are reproducible, the paper makes a useful empirical contribution: it shows that a lightweight, modular loop of specialized models can outperform much larger foundation models on crowded multi-body scenes, and that mask conditioning helps top-down pose estimation. The evaluation protocol is a strength: no OCHuman training, identical hyperparameters on COCO and OCHuman, and a component-wise ablation. The paper also ships a candid failure analysis. However, the central gain mechanism (mask-out re-detection in Eq. 7) is not quantitatively characterized, the COCO 'SOTA' claim is overreaching as stated, and the main ablation table is presented in a way that is difficult to verify.

major comments (4)
  1. [Eq. (7), Sec. 3.1; Sec. 5; Supp. Sec. C] The mask-out step in Eq. (7) is the key to the re-detection gains reported in Tables 1 and 2, but its robustness is not quantified. The paper's own limitations paragraph and Supplement Sec. C state that an over-expanded SAM mask can permanently hide a second person, that under-expanded masks leave clothing or limbs that are re-detected as false instances, and that the P-Mc filter (Eq. 6) returns 1.0 for a mask that merges instances because no negative keypoints fall outside it. Since the second iteration accounts for a large part of the OCHuman improvement (Tab. 1: 46.6 to 49.2 pose AP), the central claim depends on an unmeasured property of SAM mask accuracy. Please add a sensitivity analysis (e.g., dilating or eroding masks before Eq. 7, or injecting known over/under-coverage) and report mask-quality statistics such as mask IoU against ground truth on OCHuman. This is a load-bearing point, not a stylistic request.
  2. [Table 4 (and Supp. Table 10)] The ablation table is difficult to parse: the rows contain '1×/2×' tokens that do not correspond to a column header, and the meaning of the 'loops' column is ambiguous—the first two rows both have 'loops ✗' but are called BMP 1× and BMP 2× in the caption. The text in Sec. 4.3 refers to rows by position (e.g., 'third row'), but without a clear column mapping the reader cannot verify which configuration produced which number. Because the ablation is the evidence for the claim that each element of the loop contributes, please reformat the table with explicit columns for detector, MaskPose, SAM, mask-out re-detection, number of iterations, bbox AP, and pose AP, and make each row's configuration explicit.
  3. [Abstract and Sec. 4.2 (Table 1)] The claim that BMP 'achieves SOTA performance on COCO pose estimation' is not supported by the comparison in Table 1. The reported COCO val AP is 76.5, only 0.1 above ViTPose-B, and the table does not include larger ViTPose backbones or other leading COCO pose models, which achieve higher AP. Please qualify the claim (for example, 'SOTA among top-down models of comparable size') or expand the comparison; as written, the abstract overstates the contribution.
  4. [Abstract and Table 3] The abstract's statement that BMP 'improves detection by 39% over the baseline detector' is not reproducible from Table 3. In the 0.6–0.8 max-IoU bin, the relative improvement is (21.5 − 15.7)/15.7 ≈ 37%, and other bins show smaller relative gains. Please specify the exact subset and metric used for the 39% figure, or correct the number.
minor comments (6)
  1. [Eq. (6)] The notation is inconsistent: the text defines k_p and k_n but then uses '|k^-_p|' for the negative-keypoint term; please fix the subscripts.
  2. [Algorithm 1] 'Set ot detected keypoints' should read 'Set of detected keypoints', and 'furthest to K_s' should be 'furthest from K_s'.
  3. [Sec. 4.2] In the paragraph on BMP 2×, the text says 'from 46.6 to 49.3 AP' but Table 1 reports 49.2; please align the number.
  4. [Sec. 4.1] The term 'COCO-human' is undefined; please specify whether this is the COCO person subset and which image split is used for fine-tuning.
  5. [Sec. 4.2] 'BUTCD' is a typo for 'BUCTD' in the paragraph comparing with the state of the art.
  6. [Table 3] The column header 'bbox AP @ max IoU 0.0 – 0.2' etc. should clarify that these are range bins, since the current header could be misread as a single IoU list.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper reports empirical benchmark evaluations whose components are independently trained on external data; no derivation reduces to a fitted parameter or to a self-citation.

full rationale

The paper's central claims are empirical results on held-out test sets (OCHuman test, COCO val), not analytical derivations, so most circularity patterns do not apply. None of the models in the loop were trained on OCHuman, and the reported numbers are test-set AP values from Tables 1 and 2. The mask-out step in Eq. (7) is not circular: the detector is fine-tuned with instance-removal augmentation on COCO-human, and re-detection on masked images is an independently evaluated behavior, not a quantity defined by the mask. MaskPose (Eq. (4)) is trained on external keypoint annotations (COCO, MPII, AIC) with pseudo masks generated by SAM2, but the pose targets are ground-truth keypoints, so the pose result is not defined by the masks. The P-Mc filter (Eq. (6)) is a consistency check that discards some SAM-refined masks; it does not force the final benchmark numbers because it only selects among masks, and the paper explicitly shows it can fail when masks merge instances (Supplement C). Hyperparameters such as Nmax, Tc, and alpha were tuned on OCHuman-val, but the headline results are test AP, and tuning validation hyperparameters is a standard evaluation practice rather than a circular derivation. The paper contains no load-bearing self-citations or imported uniqueness theorems; its cited components (RTMDet, ViTPose, SAM2, BUCTD) are external prior work. The limitations section candidly identifies imperfect SAM mask refinement as the main source of errors, but this is a robustness concern, not evidence that the reported improvements reduce to the method's own inputs. Overall, the derivation chain is self-contained as an empirical system: each component is trained on external data and evaluated on a test set the method never trained on, and no prediction is constructed from the quantity it claims to predict.

Assumptions & free parameters 3 free parameters · 3 assumptions · 0 invented entities

No invented physical entities. The paper introduces a new model architecture and an iterative loop, not new particles, forces, dimensions, or conserved quantities. The main external inputs are pretrained models and public datasets.

free parameters (3)
  • alpha (mask transparency in MaskPose, Eq. 4) = 0.8
    Chosen after preliminary experiments on OCHuman where alpha in [0.2, 0.8] gave similar performance; not derived from first principles.
  • Nmax (maximum number of SAM keypoint prompts) = 6
    Selected from OCHuman-val ablations in Tables 6 to 8 as a middle ground between detection and segmentation performance.
  • Tc (SAM prompt confidence threshold) = 0.5
    Selected from OCHuman-val ablation; Table 8 shows lower thresholds gave better numbers in the ablation, yet the final method uses 0.5.
assumptions (3)
  • domain assumption SAM2-generated pseudo-ground-truth masks for MPII and AIC are accurate enough to train MaskPose.
    Section 4.1 states pseudo masks are generated by SAM2 prompted with ground-truth boxes and visible keypoints; no validation of mask quality is reported.
  • domain assumption Instance masks produced by RTMDet and SAM in the loop are accurate enough that non-transparent masking out in Eq. 7 removes one person without removing or splitting others.
    Sections 3.1 and 5 acknowledge incorrect masks cause information loss and missed re-detections; the loop's gains depend on this assumption.
  • domain assumption Keypoint confidence is a usable proxy for visibility when selecting SAM prompts.
    Section 3.3 approximates visibility by confidence because the authors could not train a reliable visibility predictor.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Detection, Pose Estimation and Segmentation for Multiple Bodies: Closing the Virtuous Circle." pith.science (2026). https://pith.science/paper/LFZLH6PT

@misc{pith2026241201562,
  author       = {Pith},
  title        = {Pith review of: Detection, Pose Estimation and Segmentation for Multiple Bodies: Closing the Virtuous Circle},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LFZLH6PT}},
  note         = {Machine review of arXiv:2412.01562}
}
read the original abstract

Human pose estimation methods work well on isolated people but struggle with multiple-bodies-in-proximity scenarios. Previous work has addressed this problem by conditioning pose estimation by detected bounding boxes or keypoints, but overlooked instance masks. We propose to iteratively enforce mutual consistency of bounding boxes, instance masks, and poses. The introduced BBox-Mask-Pose (BMP) method uses three specialized models that improve each other's output in a closed loop. All models are adapted for mutual conditioning, which improves robustness in multi-body scenes. MaskPose, a new mask-conditioned pose estimation model, is the best among top-down approaches on OCHuman. BBox-Mask-Pose pushes SOTA on OCHuman dataset in all three tasks - detection, instance segmentation, and pose estimation. It also achieves SOTA performance on COCO pose estimation. The method is especially good in scenes with large instances overlap, where it improves detection by 39% over the baseline detector. With small specialized models and faster runtime, BMP is an effective alternative to large human-centered foundational models. Code and models are available on https://MiraPurkrabek.github.io/BBox-Mask-Pose.

Figures

Figures reproduced from arXiv: 2412.01562 by the authors.

Figure 1
Figure 1. The BBox-Mask-Pose (BMP) method. Steps (A) – (D) repeat until no new detections found in step (A). Here, the background player is undetected in step (A1). BMP correctly fits the foreground player’s pose (B1) which leads to correction of his segmentation and bbox (C1). After masking the foreground player (D1), the background player is detected (A2), his body correctly segemented and pose estimated. Right: the BMP out… view at source ↗
Figure 2
Figure 2. BMP resolves detection errors (top and middle) and pose errors (bottom) on OCHuman. Quantitative results in Tab. 3. 3 [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. SAM: influence of prompting parameters. MaskPose has approximately the same number of pa￾rameters as ViTPose, differing only in head architecture and preprocessing. These small changes enable MaskPose to perform similarly on standard datasets (COCO, AIC, MPII) while improving performance in multi-body scenar￾ios. Mask conditioning adapts the top-down method for multi-body cases, allowing detailed instance specificat… view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: Typical errors of the BMP loop. The weakest part is SAM and its prompting with correct keypoints. ing falls short compared to human interaction and most of the errors come from incorrect masks. 6. The modular structure of BMP enables further perfor￾mance gains by integ…
Figure 5
Figure 5. Figure 5: Segmentation error involving a small number of pixels, [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]
Figure 6
Figure 6. Figure 6: Multiple background instances may merge into a single [PITH_FULL_IMAGE:figures/full_fig_p014_6.png]
Figure 7
Figure 7. Figure 7: Instances not split even after mask refinement by SAM [PITH_FULL_IMAGE:figures/full_fig_p015_7.png]
Figure 8
Figure 8. Figure 8: Oversegmentation. Green instances have incorrect [PITH_FULL_IMAGE:figures/full_fig_p015_8.png]
Figure 10
Figure 10. Figure 10: MaskPose performance with different values of α. Fine-tuning for 5 epochs on 10% of dataset, masks detected. pose SAM pose loops bbox pose params ✓ ✓ ✗ 1× 31.1 45.3 225 M ✓ ✓ ✗ 2× 32.1 48.6 369 M ✓ ✓ ✓ 1× 31.1 46.4 312 M ✗ ✓ ✗ 2× 31.9 47.3 282 M ✓ ✗ ✗ 2× 30.8 47.0 201…
Figure 11
Figure 11. Figure 11: Images where BMP improves detection and segmen [PITH_FULL_IMAGE:figures/full_fig_p016_11.png]
Figure 12
Figure 12. Figure 12: Two iterations of BMP successfully decouple merged [PITH_FULL_IMAGE:figures/full_fig_p017_12.png]
Figure 14
Figure 14. Figure 14: More qualitative results on the OCHuman dataset. [PITH_FULL_IMAGE:figures/full_fig_p018_14.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 25 canonical work pages

  1. [1]

    Multiposeseg: Feedback knowledge transfer for multi-person pose estimation and instance segmentation

    Niaz Ahmad, Jawad Khan, Jeremy Yuhyun Kim, and Young- moon Lee. Multiposeseg: Feedback knowledge transfer for multi-person pose estimation and instance segmentation. 2022 26th International Conference on Pattern Recognition (ICPR), pages 2086–2092, 2022. 3

    work page 2022

  2. [2]

    Joint human pose estimation and instance seg- mentation with poseplusseg

    Niaz Ahmad, Jawad Khan, Jeremy Yuhyun Kim, and Young- moon Lee. Joint human pose estimation and instance seg- mentation with poseplusseg. In AAAI Conference on Artifi- cial Intelligence, 2022. 3

    work page 2022

  3. [3]

    2d human pose estimation: New benchmark and state of the art analysis

    Mykhaylo Andriluka, Leonid Pishchulin, Peter Gehler, and Bernt Schiele. 2d human pose estimation: New benchmark and state of the art analysis. In IEEE Conference on Com- puter Vision and Pattern Recognition (CVPR), 2014. 2

    2014

  4. [4]

    k-means++: The advantages of careful seeding

    David Arthur and Sergei Vassilvitskii. k-means++: The advantages of careful seeding. Technical report, Stanford,

  5. [5]

    Test-time adaptation vs

    Kambiz Azarian, Debasmit Das, Hyojin Park, and Fatih Mu- rat Porikli. Test-time adaptation vs. training-time general- ization: A case study in human instance segmentation using keypoints estimation. 2023 IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW) , pages 411–420, 2022. 3

    work page 2023

  6. [6]

    Crowd-sam: Sam as a smart annotator for object de- tection in crowded scenes

    Zhi Cai, Yingjie Gao, Yaoyan Zheng, Nan Zhou, and Di Huang. Crowd-sam: Sam as a smart annotator for object de- tection in crowded scenes. In Proceedings of the European Conference on Computer Vision (ECCV), 2024. 6

    work page 2024

  7. [7]

    Realtime multi-person 2d pose estimation using part affinity fields

    Zhe Cao, Tomas Simon, Shih-En Wei, and Yaser Sheikh. Realtime multi-person 2d pose estimation using part affinity fields. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 7291–7299, 2017. 1, 2

    2017

  8. [8]

    Unihcp: A unified model for human-centric perceptions

    Yuanzheng Ci, Yizhou Wang, Meilin Chen, Shixiang Tang, Lei Bai, Feng Zhu, Rui Zhao, Fengwei Yu, Donglian Qi, and Wanli Ouyang. Unihcp: A unified model for human-centric perceptions. 2023 IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition (CVPR), pages 17840–17852,

    work page 2023

Show all 43 references
  1. [9]

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

    Alexey Dosovitskiy. An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929, 2020. 4

    2010 arXiv

  2. [10]

    Bottom-up human pose estimation via disentan- gled keypoint regression

    Zigang Geng, Ke Sun, Bin Xiao, Zhaoxiang Zhang, and Jing- dong Wang. Bottom-up human pose estimation via disentan- gled keypoint regression. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 14676–14686, 2021. 2, 6

    2021

  3. [11]

    Deepsportlab: a uni- fied framework for ball detection, player instance segmen- tation and pose estimation in team sports scenes

    Seyed Abolfazl Ghasemzadeh, Gabriel Van Zandycke, Maxime Istasse, Niels Sayez, Amirafshar Moshtaghpour, and Christophe De Vleeschouwer. Deepsportlab: a uni- fied framework for ball detection, player instance segmen- tation and pose estimation in team sports scenes. ArXiv, abs/...

    2021 arXiv

  4. [12]

    Instance-level human parsing via part grouping network

    Ke Gong, Xiaodan Liang, Yicheng Li, Yimin Chen, Ming Yang, and Liang Lin. Instance-level human parsing via part grouping network. In Proceedings of the European confer- ence on computer vision (ECCV) , pages 770–785, 2018. 2, 7, 14

    2018

  5. [13]

    On the calibration of human pose estimation

    Kerui Gu, Rongyu Chen, and Angela Yao. On the calibration of human pose estimation. arXiv preprint arXiv:2311.17105,

    arXiv

  6. [14]

    You only learn one query: learning unified human query for single-stage multi-person multi-task human-centric perception

    Sheng Jin, Shuhuai Li, Tong Li, Wentao Liu, Chen Qian, and Ping Luo. You only learn one query: learning unified human query for single-stage multi-person multi-task human-centric perception. In European Conference on Computer Vision , pages 126–146. Springer, 2024. 6

    2024

  7. [15]

    Multi-instance pose networks: Rethinking top-down pose estimation

    Rawal Khirodkar, Visesh Chari, Amit Agrawal, and Ambrish Tyagi. Multi-instance pose networks: Rethinking top-down pose estimation. 2021 IEEE/CVF International Conference on Computer Vision (ICCV), pages 3102–3111, 2021. 1, 2, 6

    2021

  8. [16]

    Sapiens: Foundation for human vision models

    Rawal Khirodkar, Timur Bagautdinov, Julieta Martinez, Zhaoen Su, Austin James, Peter Selednik, Stuart Anderson, and Shunsuke Saito. Sapiens: Foundation for human vision models. In European Conference on Computer Vision, 2024. 2, 6, 7

    2024

  9. [17]

    Crowdpose: Efficient crowded scenes pose estimation and a new benchmark

    Jiefeng Li, Can Wang, Hao Zhu, Yihuan Mao, Hao-Shu Fang, and Cewu Lu. Crowdpose: Efficient crowded scenes pose estimation and a new benchmark. arXiv preprint arXiv:1812.00324, 2018. 2

    2018 arXiv

  10. [18]

    Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Doll ´ar, and C

    Tsung-Yi Lin, Michael Maire, Serge J. Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Doll ´ar, and C. Lawrence Zitnick. Microsoft coco: Common objects in context. In European Conference on Computer Vision, 2014. 2

    2014

  11. [19]

    Humans need not label more humans: Occlusion copy & paste for occluded human instance segmentation

    Evan Ling, De-Kai Huang, and Minhoe Hur. Humans need not label more humans: Occlusion copy & paste for occluded human instance segmentation. In British Machine Vision Conference, 2022. 6

    2022

  12. [20]

    Swin transformer: Hierarchical vision transformer using shifted windows

    Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, and Baining Guo. Swin transformer: Hierarchical vision transformer using shifted windows. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 10012–10022, 2021. 2

    2021

  13. [21]

    A convnet for the 2020s

    Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feicht- enhofer, Trevor Darrell, and Saining Xie. A convnet for the 2020s. Proceedings of the IEEE/CVF Conference on Com- puter Vision and Pattern Recognition (CVPR), 2022. 3, 6

    2022

  14. [22]

    Rtmdet: An empirical study of designing real-time object detectors

    Chengqi Lyu, Wenwei Zhang, Haian Huang, Yue Zhou, Yudong Wang, Yanyi Liu, Shilong Zhang, and Kai Chen. Rtmdet: An empirical study of designing real-time object detectors. ArXiv, abs/2212.07784, 2022. 2, 3, 4, 6, 11, 12, 13, 14, 16, 17, 18

    2022 arXiv

  15. [23]

    George Papandreou, Tyler Lixuan Zhu, Nori Kanazawa, Alexander Toshev, Jonathan Tompson, Christoph Bregler, and Kevin P. Murphy. Towards accurate multi-person pose estimation in the wild. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , pages 3711–3719,

    2017

  16. [24]

    George Papandreou, Tyler Lixuan Zhu, Liang-Chieh Chen, Spyros Gidaris, Jonathan Tompson, and Kevin P. Murphy. Personlab: Person pose estimation and instance segmen- tation with a bottom-up, part-based, geometric embedding model. In European Conference on Computer Vision, 2018. 2

    2018

  17. [25]

    Sam 2: Segment anything in images and videos

    Nikhila Ravi, Valentin Gabeur, Yuan-Ting Hu, Ronghang Hu, Chaitanya Ryali, Tengyu Ma, Haitham Khedr, Roman 9 R¨adle, Chloe Rolland, Laura Gustafson, et al. Sam 2: Segment anything in images and videos. arXiv preprint arXiv:2408.00714, 2024. 2, 3, 4, 11, 12, 13, 15

    2024 arXiv

  18. [26]

    Iterdet: iterative scheme for object detection in crowded environments

    Danila Rukhovich, Konstantin Sofiiuk, Danil Galeev, Olga Barinova, and Anton Konushin. Iterdet: iterative scheme for object detection in crowded environments. In Structural, syntactic, and statistical pattern recognition: Joint IAPR in- ternational workshops, s+ SSPR 2020, pad...

    2020

  19. [27]

    Crowdhuman: A benchmark for detecting human in a crowd

    Shuai Shao, Zijian Zhao, Boxun Li, Tete Xiao, Gang Yu, Xi- angyu Zhang, and Jian Sun. Crowdhuman: A benchmark for detecting human in a crowd. ArXiv, abs/1805.00123, 2018. 2

    2018 arXiv

  20. [28]

    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

    2022

  21. [29]

    End-to- end trainable multi-instance pose estimation with transform- ers

    Lucas Stoffl, Maxime Vidal, and Alexander Mathis. End-to- end trainable multi-instance pose estimation with transform- ers. arXiv preprint arXiv:2103.12115, 2021. 2

    2021 arXiv

  22. [30]

    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. 2, 3, 6

    2019

  23. [31]

    Rethinking visibility in human pose esti- mation: Occluded pose reasoning via transformers

    Pengzhan Sun, Kerui Gu, Yunsong Wang, Linlin Yang, and Angela Yao. Rethinking visibility in human pose esti- mation: Occluded pose reasoning via transformers. 2024 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), pages 5891–5900, 2024. 5

    2024

  24. [32]

    Collins, Matthew A

    Subarna Tripathi, Maxwell D. Collins, Matthew A. Brown, and Serge J. Belongie. Pose2instance: Harnessing keypoints for person instance segmentation. ArXiv, abs/1704.01152,

    arXiv

  25. [33]

    Contextual instance decoupling for robust multi-person pose estimation

    Dongkai Wang and Shiliang Zhang. Contextual instance decoupling for robust multi-person pose estimation. 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 11050–11058, 2022. 1, 2, 6

    2022

  26. [34]

    Internimage: Exploring large-scale vi- sion foundation models with deformable convolutions

    Wenhai Wang, Jifeng Dai, Zhe Chen, Zhenhang Huang, Zhiqi Li, Xizhou Zhu, Xiaowei Hu, Tong Lu, Lewei Lu, Hongsheng Li, et al. Internimage: Exploring large-scale vi- sion foundation models with deformable convolutions. In Proceedings of the IEEE/CVF conference on computer vi- si...

    2023

  27. [35]

    Hulk: A universal knowledge translator for human-centric tasks

    Yizhou Wang, Yixuan Wu, Shixiang Tang, Weizhen He, Xun Guo, Feng Zhu, Lei Bai, Rui Zhao, Jian Wu, Tong He, and Wanli Ouyang. Hulk: A universal knowledge translator for human-centric tasks. ArXiv, abs/2312.01697, 2023. 2

    2023 arXiv

  28. [36]

    Ai challenger : A large-scale dataset for going deeper in image understanding

    Jiahong Wu, He Zheng, Bo Zhao, Yixin Li, Baoming Yan, Rui Liang, Wenjia Wang, Shipei Zhou, Guosen Lin, Yanwei Fu, Yizhou Wang, and Yonggang Wang. Ai challenger : A large-scale dataset for going deeper in image understanding. ArXiv, abs/1711.06475, 2017. 2

    2017 arXiv

  29. [37]

    Adaptivepose++: A powerful single-stage network for multi-person pose re- gression

    Yabo Xiao, Xiaojuan Wang, Dongdong Yu, Kai Su, Lei Jin, Mei Song, Shuicheng Yan, and Jian Zhao. Adaptivepose++: A powerful single-stage network for multi-person pose re- gression. arXiv preprint arXiv:2210.04014, 2022. 2

    2022 arXiv

  30. [38]

    ViTPose: Simple vision transformer baselines for human pose estimation

    Yufei Xu, Jing Zhang, Qiming Zhang, and Dacheng Tao. ViTPose: Simple vision transformer baselines for human pose estimation. In Advances in Neural Information Pro- cessing Systems, 2022. 1, 2, 3, 4, 6, 14

    2022

  31. [39]

    Hrformer: High- resolution vision transformer for dense predict

    Yuhui Yuan, Rao Fu, Lang Huang, Weihong Lin, Chao Zhang, Xilin Chen, and Jingdong Wang. Hrformer: High- resolution vision transformer for dense predict. Advances in neural information processing systems , 34:7281–7293,

  32. [40]

    Pose2seg: Detection free human instance segmentation

    Song-Hai Zhang, Ruilong Li, Xin Dong, Paul Rosin, Zixi Cai, Xi Han, Dingcheng Yang, Haozhi Huang, and Shi-Min Hu. Pose2seg: Detection free human instance segmentation. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 889–898, 2019. 2, 3

    2019

  33. [41]

    Poseg: Pose-aware refinement network for human instance segmentation

    Desen Zhou and Qian He. Poseg: Pose-aware refinement network for human instance segmentation. IEEE Access, 8: 15007–15016, 2020. 3, 6, 7

    2020

  34. [42]

    Mathis, and Alexan- der Mathis

    Mu Zhou, Lucas Stoffl, Mackenzie W. Mathis, and Alexan- der Mathis. Rethinking pose estimation in crowds: over- coming the detection information bottleneck and ambiguity. 2023 IEEE/CVF International Conference on Computer Vi- sion (ICCV), pages 14643–14653, 2023. 1, 2, 6, 7, 8

    2023

  35. [43]

    Detrs with col- laborative hybrid assignments training

    Zhuofan Zong, Guanglu Song, and Yu Liu. Detrs with col- laborative hybrid assignments training. In Proceedings of the IEEE/CVF international conference on computer vision, pages 6748–6758, 2023. 2, 6 10 Detection, Pose Estimation and Segmentation for Multiple Bodies: Closing t...

    2023

Pith tools

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