Pith. sign in

REVIEW 3 major objections 5 minor 38 references

Category Level 6D Object Pose Estimation from a Single RGB Image using Diffusion

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read A score-based diffusion model that generates pose hypotheses and picks the mode with Mean Shift achieves state-of-the-art RGB-only category-level 6D pose estimation on REAL275.

desk verdict A credible diffusion-based pose estimator with a neat mean-shift trick, but the headline SOTA gain is tied to an ablation row that doesn't match the described architecture. read the letter →

arxiv 2412.11420 v1 pith:YRCKCSPA submitted 2024-12-16 cs.CV

Adam Bethell , Ravi Garg , Ian Reid This is my paper · ORCID
classification cs.CV
keywords category-level6DobjectposeestimationsingleRGBimagescore-baseddiffusionmodelsmeanshiftmodeseekingREAL275benchmarkhypothesisdistributionscale-awareNOCS
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

This paper aims to solve category-level 6D object pose and size estimation from a single RGB image, without CAD models or depth sensors. The authors argue that pose estimation is naturally a distribution over hypotheses, because symmetries, occlusions, and scale ambiguity make several poses plausible for the same observation. They build a score-based diffusion model conditioned on the RGB crop, predicted relative depth and normals, object category, and a global image feature to generate a set of pose hypotheses, then use Mean Shift to select the mode of this distribution as the final pose. On the REAL275 benchmark the method reports state-of-the-art results for RGB-based methods, with a 55% relative improvement on the strict 10°10cm metric and an 18% improvement on IoU75 over the previous best (DMSR). The practical payoff is that accurate category-level pose estimation becomes possible with a single ordinary camera.

What carries the argument

The central object is a score-based diffusion model (ScoreNet) that learns the score — the gradient of the log-density of the pose given the observation — under the Variance-Exploding SDE. Pose is parameterised as a 12D vector combining a continuous 6D rotation, scale-invariant translation parameters (SITE), and 3D size, and the network is conditioned on semantic features from the RGB crop, geometric features from predicted relative depth and normals, a learned category-ID embedding, and a global image feature that preserves scene context lost by cropping. At inference the Probability Flow ODE denoises Gaussian noise into K pose hypotheses that approximate the conditional pose distribution, and Mean Shift++ is applied separately to rotation, translation, and size to return the mode of the distribution as the final pose. This replaces the two-stage procedure of GenPose, which trains a separate energy-based likelihood model to discard outliers before mean pooling, with a single mode-seeking step.

What would settle it

Run REAL275 evaluation with the Omnidata DPT depth/normal features replaced by an alternative predictor (or by corrupted depth/normals) and check whether the 10°10cm accuracy collapses toward the GenPose-with-UniDepth baseline; or take a category with high intra-class scale variation (e.g., chairs) and test whether the category-ID scale prior alone yields accurate size estimates.

Watch

Extended reading notes

Core claim

The central claim is that modelling the pose posterior with a conditional score-based diffusion model, and reading out the mode of the sampled hypotheses with Mean Shift, is enough to beat prior RGB-only category-level pose estimators by a wide margin. The ScoreNet is trained with the denoising score-matching objective of the variance-exploding SDE, conditioned on two CNN encoders: one for the RGB crop and one for concatenated relative depth and normal maps predicted by a pretrained Omnidata DPT model, plus a learned category embedding and a global feature from the 2D detector. At inference, the Probability Flow ODE draws K=50 pose hypotheses, and Mean Shift++ finds the mode of the rotation, translation, and size distributions. On REAL275 the method reports 36.6 accuracy on the 10°10cm metric vs 23.6 for DMSR, and 45.2 vs 28.3 on IoU50, while also being simpler than the prior approach that trains a separate energy model to filter hypotheses before mean pooling. The paper further shows that replacing the point-cloud conditioning of GenPose with back-projected metric depth from UniDepth performs much worse, and that the category embedding and global feature carry the scale information that metric depth would otherwise provide.

Load-bearing premise

The method's accuracy depends on the predicted relative depth and normal maps being geometrically trustworthy enough for the ScoreNet to infer pose and scale; if that predictor fails, the pose estimates degrade, as the paper acknowledges in its conclusion.

Editorial extensions

If this is right

  • RGB-only category-level pose estimation can reach accuracy previously requiring depth sensors, since the geometric cues come from predicted relative depth and normals rather than measured point clouds.
  • The extra energy/likelihood network used by prior generative pose methods is unnecessary: Mean Shift on the diffusion samples yields the final pose at higher accuracy and roughly twice the speed of the energy-model pipeline.
  • The same ScoreNet supports tracking: initialising the hypotheses of the next frame from the current frame's hypotheses gives a category-level RGB pose tracker with minimal modification.
  • Scale estimation can be conditioned on object category and global image features, so metric depth prediction — which is scale-error prone — is not needed for the pose pipeline.
  • The gap between the Mean Shift result (36.6) and the oracle best-hypothesis result (83.0) on 10°10cm indicates the ScoreNet's distribution is highly accurate, and better mode-finding or hypothesis selection could give further gains.

Reading between the lines

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

  • The method inherits the failure modes of the pretrained depth-and-normal predictor: the reported gains depend on Omnidata DPT being a reliable source of geometric features, so categories or environments where those predictions are poor will likely show degraded translation and size accuracy, as the paper itself concedes.
  • Choosing the mode rather than the mean is principled for peaked, unimodal pose posteriors, but for perfectly symmetric objects the conditional distribution is uniform along the symmetry axis, so the 'mode' is arbitrary; the paper's evaluation partially finesses this by ignoring rotation error around the symmetry axis.
  • Running Mean Shift independently on rotation, translation, and size ignores correlations between these components; a joint mode-seeking procedure on the full pose manifold could be a testable extension.
  • Because the global image feature carries scene context used for scale, the method's behaviour on novel object categories not seen in training is untested; the category-dropout mechanism suggests it should degrade gracefully, but that is an inference, not a demonstrated result.
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 / 5 minor

Summary. The paper proposes a score-based diffusion model for category-level 6D object pose and size estimation from a single RGB image. The method conditions a ScoreNet on RGB crops, predicted relative depth and normal maps from a pretrained Omnidata DPT model, category IDs, and a global image feature, then generates K pose hypotheses and selects the final pose by running Mean Shift on the rotation, translation, and size components. The authors report state-of-the-art results on REAL275, together with ablations on the conditioning features, the number of hypotheses, the aggregation strategy, and a tracking extension.

Significance. If the results are correct as presented, the paper would be a meaningful advance: it is one of the first diffusion-based frameworks for RGB-only category-level pose estimation, it avoids the category shape prior used by DMSR, and it replaces GenPose's expensive energy-model filtering with a mode-seeking step. The ablation study on the number of pose hypotheses and the comparison with a metric-depth-adapted GenPose are useful and informative. However, the central SOTA claim is currently tied to a table row that does not match the architecture described in the method section, and the advantage of Mean Shift over plain mean pooling is not demonstrated by the reported numbers. These issues must be resolved before the claims can be accepted.

major comments (3)
  1. [§1, §4.3, Table 1] The headline claims in the Abstract and §1 -- a 55% improvement on the strict 10°10cm metric and 18% on IoU75 -- are computed from the 'Ours - NGR' row of Table 1 (36.6 vs. 23.6 and 7.2 vs. 6.1), not from the 'Ours' row (32.7 and 6.7). The method described in §3.2 and Figure 5 concatenates the global feature G with the other conditioning embeddings and feeds the resulting vector to all four decoder heads, with no mention of withholding G from the rotation head; 'Ours - NGR' is defined in §4.3 as providing the global feature only to the translation and size heads. Tables 2, 3, and 4 all report the 36.6-level numbers, so the experimental core is built on the NGR configuration while the architecture section describes the Ours configuration. Please either identify NGR as the actual method and revise §3.2 and Figure 5 accordingly, or recompute all headline gains from the Ours row (the gain on 10°10cm would be about 38.6%, not 55%). The same inconsistency affects the '14% and 52%' rotation/translation increases in §4.3, which correspond to NGR (67.5 vs. 59.5; 56.6 vs. 37.3) rather than Ours (61.3; 56.1).
  2. [§4.4.3, Table 4] The paper's second main contribution is that Mean Shift replaces the trained energy model and mean pooling, but Table 4 shows the 'None' mean-pooling baseline achieves 36.4 on the 10°10cm metric at 6.7 FPS, essentially identical to Mean Shift's 36.6 at 6.8 FPS. Since mean pooling is simpler and equally fast, the claimed advantage over the mean-pooling baseline is negligible. The comparison against the energy model (31.5, 3.6 FPS) is favorable, but mean pooling also beats that baseline. Please provide evidence that Mean Shift is preferable to mean pooling in multimodal cases, or temper the contribution claim accordingly.
  3. [§4.3, Table 1 (CAMERA columns)] On the CAMERA split, the reported Ours/Ours-NGR results are substantially worse than DMSR on IoU50 (22.9/22.9 vs. 34.6), IoU75 (3.2/3.1 vs. 6.5), 10cm (22.4/22.1 vs. 32.3), and 10°10cm (19.2/18.9 vs. 27.4), while the text only mentions the 5% improvement on the rotation metric. Please report and discuss the CAMERA shortfall rather than highlighting only the metric where the method wins.
minor comments (5)
  1. [Tables 1 and 3] The column headers '10o 10cm10 o10cm' and '10o' are garbled; use '10°', '10cm', and '10°10cm' with clear separators, and define the 5o2cm metric mentioned in §4.4.3.
  2. [§3.1] Typo: 'MaskRCCN' should be 'Mask R-CNN'.
  3. [§3.3] The phrase 'classifier-free guidance' is used for random condition dropping, but no guidance scale is applied at inference; rename this 'condition dropout' or define the guidance procedure explicitly.
  4. [§4.5, Table 5] The tracking comparison should state the input modality of the GenPose baseline (depth/point cloud vs. RGB); as presented, the table is difficult to interpret and the label 'strong baseline' is not supported by the reported numbers.
  5. [Eq. (1)] The notation p(0) and p(t) is not defined in the main text; define these as the clean and perturbed pose samples respectively.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the pose estimate is produced by a learned ScoreNet plus Mean Shift, and the claimed gains are empirical comparisons; the reported SOTA number comes from an ablation row rather than the described configuration, which is a reporting inconsistency and not a circular reduction.

full rationale

The paper's derivation chain is concrete and empirical: (i) a ScoreNet adapted from GenPose is conditioned on RGB, predicted relative depth/normals, category ID, and a global image feature; (ii) it is trained with the score-matching loss in Eq. (1); (iii) at inference K pose hypotheses are sampled via the Probability Flow ODE; (iv) Mean Shift is run on the hypotheses to select the mode as the final pose. None of these steps defines a prediction in terms of its own target. The category-ID and global-feature inputs come from MaskRCNN pre-processing; they are not fitted to the pose labels in a way that would force the final pose. The ablations Ours-NG, Ours-NC, and Ours-NGR vary conditioning information and are therefore legitimate analysis, not a fitted-parameter-renamed-as-prediction. Mean Shift is a post-hoc aggregator over stochastically generated hypotheses and does not by construction force the output to match the ground truth. The GenPose-with-UniDepth comparison is likewise empirical. The main flagged issue is that the abstract's "55% improvement" and several later tables correspond to the Ours-NGR row (36.6 on 10°10cm) while the architecture in §3.2 and Figure 1 describes the Ours row (32.7); this is a reporting/consistency defect in which an ablation variant is used for the headline comparison, but it is not a case where the claimed result reduces to its inputs by construction. There is no load-bearing self-citation: GenPose [35] and DMSR [33] are external works by different author groups, and no uniqueness theorem is imported from the authors' own prior work. The conclusion's admission that the method is "reliant on the quality of the predicted depths and normals" is a stated limitation, not a circular step. Overall, the derivation is self-contained and the circularity score is 0.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

The method does not introduce new physical entities or forces. It relies on standard pretrained models (Omnidata, MaskRCNN) and standard deep learning components. The free parameters are the usual hyperparameters and the number of hypotheses, which is tuned. The main load-bearing assumptions are the reliability of the external pretrained models and the usefulness of the global feature for scale.

free parameters (1)
  • Number of pose hypotheses K = 50
    Chosen based on an ablation (Table 3) that shows diminishing returns beyond 50. This is a hyperparameter that affects performance and is tuned on the validation set.
assumptions (3)
  • domain assumption The pretrained Omnidata DPT model provides accurate relative depth and normals from a single RGB image.
    The method's conditioning features (depths and normals) come from this model. If these are inaccurate, the pose hypotheses and final pose estimate are likely to be inaccurate, as the paper's conclusion admits.
  • domain assumption The MaskRCNN detector with COCO training and finetuning on the target dataset provides reliable object bounding boxes and global features.
    The whole pipeline starts with detection; the paper uses Dynamic Zoom In to be robust to detection errors, but the global feature is a 2048-dim vector from MaskRCNN, and the category ID is essential for scale. If detection is poor, the inputs degrade.
  • domain assumption A single global image feature can help resolve scale ambiguity for category-level objects.
    The paper uses the global feature to inform scale and argues it provides context. This is an empirical assumption that is not theoretically justified, and the ablation -NG shows it matters for translation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Category Level 6D Object Pose Estimation from a Single RGB Image using Diffusion." pith.science (2026). https://pith.science/paper/YRCKCSPA

@misc{pith2026241211420,
  author       = {Pith},
  title        = {Pith review of: Category Level 6D Object Pose Estimation from a Single RGB Image using Diffusion},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YRCKCSPA}},
  note         = {Machine review of arXiv:2412.11420}
}
read the original abstract

Estimating the 6D pose and 3D size of an object from an image is a fundamental task in computer vision. Most current approaches are restricted to specific instances with known models or require ground truth depth information or point cloud captures from LIDAR. We tackle the harder problem of pose estimation for category-level objects from a single RGB image. We propose a novel solution that eliminates the need for specific object models or depth information. Our method utilises score-based diffusion models to generate object pose hypotheses to model the distribution of possible poses for the object. Unlike previous methods that rely on costly trained likelihood estimators to remove outliers before pose aggregation using mean pooling, we introduce a simpler approach using Mean Shift to estimate the mode of the distribution as the final pose estimate. Our approach outperforms the current state-of-the-art on the REAL275 dataset by a significant margin.

Figures

Figures reproduced from arXiv: 2412.11420 by the authors.

Figure 1
Figure 1. An overview of our method. Preprocessing: The input image is passed through a depth and normal predictor and a 2D object detector to get the cropped images, global image feature G and category ID c. Hypothesis Generation: Pose hypotheses are initialised using Gaussian Noise and pass through the Probability Flow ODE and ScoreNet to generate the nal pose hypotheses pˆ1    pˆK. Mode Prediction: Mean Shift is then r… view at source ↗
Figure 2
Figure 2. Qualitative results of our method on the REAL275 dataset. [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. The average precision for various error thresholds for each category. [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Qualitative results of the predicted rotation distribution for a symmetrical and asymmetrical [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: The architecture of our ScoreNet. A.1 Architecture A.1.1 Dynamic Zoom In The input to our CNN encoders are cropped images of the object so following previous object pose estimation methods, we decouple the object detection and pose estimation by generating the image cr…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 34 canonical work pages

  1. [1]

    SGPA: Structure-guided prior adaptation for category-level 6d object pose estimation

    Kai Chen and Qi Dou. SGPA: Structure-guided prior adaptation for category-level 6d object pose estimation. In2021 IEEE/CVF International Conference on Computer Vision (ICCV), pages 2753–2762. ISSN: 2380-7504

    work page

  2. [2]

    Category level object pose estimation via neural analysis-by-synthesis

    Xu Chen, Zijian Dong, Jie Song, Andreas Geiger, and Otmar Hilliges. Category level object pose estimation via neural analysis-by-synthesis. In Andrea V edaldi, Horst Bischof, Thomas Brox, and Jan-Michael Frahm, editors,Computer Vision – ECCV 2020, Lecture Notes in Computer Science, pages 139–156. Springer International Publishing

    work page 2020

  3. [3]

    Comaniciu and P

    D. Comaniciu and P . Meer. Mean shift: a robust approach toward feature space analysis. 24(5):603–619

    work page

  4. [4]

    GPV-pose: Category-level object pose estimation via geometry-guided point-wise voting

    Y an Di, Ruida Zhang, Zhiqiang Lou, Fabian Manhardt, Xiangyang Ji, Nassir Navab, and Federico Tombari. GPV-pose: Category-level object pose estimation via geometry-guided point-wise voting. In2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 6771–6781. IEEE

    work page

  5. [5]

    Vision-based robotic grasping from object localization, object pose estimation to grasp estimation for parallel grippers: a review

    Guoguang Du, Kai Wang, Shiguo Lian, and Kaiyong Zhao. Vision-based robotic grasping from object localization, object pose estimation to grasp estimation for parallel grippers: a review. 54(3):1677–1734. Number: 3

    work page

  6. [6]

    Omnidata: A scalable pipeline for making multi-task mid-level vision datasets from 3d scans

    Ainaz Eftekhar, Alexander Sax, Jitendra Malik, and Amir Zamir. Omnidata: A scalable pipeline for making multi-task mid-level vision datasets from 3d scans. In2021 IEEE/CVF International Conference on Computer Vision (ICCV), pages 10766–10776. IEEE

    work page

  7. [7]

    Object level depth reconstruction for category level 6d object pose estimation from monocular RGB image

    Zhaoxin Fan, Zhenbo Song, Jian Xu, Zhicheng Wang, Kejian Wu, Hongyan Liu, and Jun He. Object level depth reconstruction for category level 6d object pose estimation from monocular RGB image. In Shai Avidan, Gabriel Brostow, Moustapha Cissé, Giovanni Maria Farinella, and Tal Hassner, editors,Computer Vision – ECCV 2022, Lecture Notes in Computer Science, p...

    work page 2022

  8. [8]

    Mask r-CNN

    Kaiming He, Georgia Gkioxari, Piotr Dollar, and Ross Girshick. Mask r-CNN. pages 2961– 2969

    work page

Show all 38 references
  1. [9]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition

  2. [10]

    Estimation of non-normalized statistical models by score matching

    Aapo Hyvarinen. Estimation of non-normalized statistical models by score matching

  3. [11]

    MeanShift++: Extremely fast mode-seeking with applications to segmentation and object tracking

    Jennifer Jang and Heinrich Jiang. MeanShift++: Extremely fast mode-seeking with applications to segmentation and object tracking. In2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 4100–4111. IEEE

  4. [12]

    3d common corruptions and data augmentation

    Oguzhan Fatih Kar, Teresa Y eo, Andrei Atanov, and Amir Zamir. 3d common corruptions and data augmentation

  5. [13]

    Category-level metric scale object shape and pose estimation

    Taeyeop Lee, Byeong-Uk Lee, Myungchul Kim, and In So Kweon. Category-level metric scale object shape and pose estimation. 6(4):8575–8582. Number: 4

  6. [14]

    CDPN: Coordinates-based disentangled pose network for real-time RGB-based 6-DoF object pose estimation

    Zhigang Li, Gu Wang, and Xiangyang Ji. CDPN: Coordinates-based disentangled pose network for real-time RGB-based 6-DoF object pose estimation. In2019 IEEE/CVF International Conference on Computer Vision (ICCV), pages 7677–7686. IEEE

  7. [15]

    SAR-net: Shape alignment and recovery network for category-level 6d object pose and size estimation

    Haitao Lin, Zichang Liu, Chilam Cheang, Y anwei Fu, Guodong Guo, and Xiangyang Xue. SAR-net: Shape alignment and recovery network for category-level 6d object pose and size estimation. In2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 6697–6707. IEEE

  8. [16]

    Category-level 6d object pose and size estimation using self-supervised deep prior deformation networks

    Jiehong Lin, Zewei Wei, Changxing Ding, and Kui Jia. Category-level 6d object pose and size estimation using self-supervised deep prior deformation networks. In Shai Avidan, Gabriel Brostow, Moustapha Cissé, Giovanni Maria Farinella, and Tal Hassner, editors,Computer Vision – ...

    2022

  9. [17]

    VI-net: Boosting category-level 6d object pose estimation via learning decoupled rotations on the spherical representations

    Jiehong Lin, Zewei Wei, Y abin Zhang, and Kui Jia. VI-net: Boosting category-level 6d object pose estimation via learning decoupled rotations on the spherical representations. pages 14001–14011

  10. [18]

    Lawrence Zitnick, and Piotr Dollár

    Tsung-Yi Lin, Michael Maire, Serge Belongie, Lubomir Bourdev, Ross Girshick, James Hays, Pietro Perona, Deva Ramanan, C. Lawrence Zitnick, and Piotr Dollár. Microsoft COCO: Common objects in context

  11. [19]

    CPS++: Improving class-level 6d pose and shape estimation from monocular images with self-supervised learning

    Fabian Manhardt, Gu Wang, Benjamin Busam, Manuel Nickel, Sven Meier, Luca Minciullo, Xiangyang Ji, and Nassir Navab. CPS++: Improving class-level 6d pose and shape estimation from monocular images with self-supervised learning

  12. [20]

    UniDepth: Universal monocular metric depth estimation

    Luigi Piccinelli, Y ung-Hsu Y ang, Christos Sakaridis, Mattia Segu, Siyuan Li, Luc V an Gool, and Fisher Y u. UniDepth: Universal monocular metric depth estimation

  13. [21]

    Berg, and Li Fei-Fei

    Olga Russakovsky, Jia Deng, Hao Su, Jonathan Krause, Sanjeev Satheesh, Sean Ma, Zhiheng Huang, Andrej Karpathy, Aditya Khosla, Michael Bernstein, Alexander C. Berg, and Li Fei-Fei. ImageNet large scale visual recognition challenge

  14. [22]

    Generative modeling by estimating gradients of the data distribution

    Y ang Song and Stefano Ermon. Generative modeling by estimating gradients of the data distribution. InAdvances in Neural Information Processing Systems, volume 32. Curran Associates, Inc

  15. [23]

    SCORE-BASED GENERA TIVE MODELING THROUGH STOCHASTIC DIFFERENTIAL EQUA TIONS

    Y ang Song, Jascha Sohl-Dickstein, Diederik P Kingma, Abhishek Kumar, Stefano Ermon, and Ben Poole. SCORE-BASED GENERA TIVE MODELING THROUGH STOCHASTIC DIFFERENTIAL EQUA TIONS

  16. [24]

    6d object pose estimation with depth images: A seamless approach for robotic interaction and augmented reality

    David Joseph Tan, Nassir Navab, and Federico Tombari. 6d object pose estimation with depth images: A seamless approach for robotic interaction and augmented reality. Issue: arXiv:1709.01459

    arXiv

  17. [25]

    3d mapping and 6d pose computation for real time augmented reality on cylindrical objects

    Fulin Tang, Yihong Wu, Xiaohui Hou, and Haibin Ling. 3d mapping and 6d pose computation for real time augmented reality on cylindrical objects. 30(9):2887–2899. Number: 9 Conference Name: IEEE Transactions on Circuits and Systems for Video Technology

  18. [26]

    Ang, and Gim Hee Lee

    Meng Tian, Marcelo H. Ang, and Gim Hee Lee. Shape prior deformation for categorical 6d object pose and size estimation. In Andrea V edaldi, Horst Bischof, Thomas Brox, and Jan-Michael Frahm, editors,Computer Vision – ECCV 2020, volume 12366, pages 530–546. Springer Internation...

    2020

  19. [27]

    Deep object pose estimation for semantic robotic grasping of household objects

    Jonathan Tremblay, Thang To, Balakumar Sundaralingam, Y u Xiang, Dieter Fox, and Stan Bircheld. Deep object pose estimation for semantic robotic grasping of household objects. Issue: arXiv:1809.10790

    arXiv

  20. [28]

    S. Umeyama. Least-squares estimation of transformation parameters between two point patterns. 13(4):376–380. Number: 4 Conference Name: IEEE Transactions on Pattern Analysis and Machine Intelligence

  21. [29]

    A connection between score matching and denoising autoencoders

    Pascal Vincent. A connection between score matching and denoising autoencoders. 23(7):1661–

  22. [30]

    GDR-net: Geometry- guided direct regression network for monocular 6d object pose estimation

    Gu Wang, Fabian Manhardt, Federico Tombari, and Xiangyang Ji. GDR-net: Geometry- guided direct regression network for monocular 6d object pose estimation. In2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 16606–16616. IEEE

  23. [31]

    He Wang, Srinath Sridhar, Jingwei Huang, Julien V alentin, Shuran Song, and Leonidas J. Guibas. Normalized object coordinate space for category-level 6d object pose and size estimation. In 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 2637–2...

    2019

  24. [32]

    Category-level 6d object pose estimation via cascaded relation and recurrent reconstruction networks

    Jiaze Wang, Kai Chen, and Qi Dou. Category-level 6d object pose estimation via cascaded relation and recurrent reconstruction networks. In2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 4807–4814. ISSN: 2153-0866. 11

  25. [33]

    RGB-based category-level object pose estimation via decoupled metric scale recovery

    Jiaxin Wei, Xibin Song, Weizhe Liu, Laurent Kneip, Hongdong Li, and Pan Ji. RGB-based category-level object pose estimation via decoupled metric scale recovery

  26. [34]

    Mean shift, mode seeking, and clustering

    Yizong Cheng. Mean shift, mode seeking, and clustering. 17(8):790–799

  27. [35]

    GenPose: Generative category-level object pose estimation via diffusion models

    Jiyao Zhang, Mingdong Wu, and Hao Dong. GenPose: Generative category-level object pose estimation via diffusion models

  28. [36]

    SSP-pose: Symmetry-aware shape prior deformation for direct category-level object pose estimation

    Ruida Zhang, Y an Di, Fabian Manhardt, Federico Tombari, and Xiangyang Ji. SSP-pose: Symmetry-aware shape prior deformation for direct category-level object pose estimation. In 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 7452–7459. IS...

    2022

  29. [37]

    On the continuity of rotation representations in neural networks

    Yi Zhou, Connelly Barnes, Jingwan Lu, Jimei Y ang, and Hao Li. On the continuity of rotation representations in neural networks. In2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 5738–5746. IEEE. 12 A Architecture and Additional Implementation...

  30. [1674]

    Number: 7 Conference Name: Neural Computation

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.