pith. sign in

arxiv: 2605.28257 · v1 · pith:KYRUPRCKnew · submitted 2026-05-27 · 💻 cs.CV

Category-Level 3D Correspondence in Camera Space via Morphable Object Priors

Leonhard Sommer , Artur Jesslen , Basavaraj Sunagad , Adam Kortylewski This is my paper

Pith reviewed 2026-06-29 13:53 UTC · model grok-4.3

classification 💻 cs.CV
keywords category-level 3D correspondencemorphable object priorscamera spaceshape disentanglementpose estimationbenchmark datasethousehold objects
0
0 comments X

The pith

Category-level 3D correspondences emerge implicitly when learning a shared morphable object prior from images.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper aims to show that predicting consistent 3D locations across different object instances in a category from a single image can happen without any direct supervision on correspondences. Instead, it arises by training a model to learn a morphable prior that separates the canonical shape of the category, how each instance deforms from it, and the pose of the object. A sympathetic reader would care because this could lead to better part-level understanding for applications like robotics without needing expensive labeled data for every correspondence. The authors introduce a new benchmark called HouseCorr3D with many images and annotations to test this. Their method called Morpheus demonstrates this by achieving strong performance on the benchmark.

Core claim

By disentangling canonical shape, deformation, and object pose in a morphable category-level shape prior learned from image reconstruction, semantically meaningful 3D correspondences in camera space emerge implicitly without explicit correspondence supervision.

What carries the argument

The morphable object prior that shares a canonical grounding across instances by disentangling shape, deformation, and pose.

If this is right

  • Semantically meaningful 3D correspondences appear in camera space across instances.
  • The approach sets a new state of the art on the HouseCorr3D benchmark.
  • Semantic 3D object understanding arises without direct correspondence supervision.
  • Consistent correspondences hold even for occluded regions due to amodal labels.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • This could allow similar implicit emergence in other vision tasks like part segmentation without supervision.
  • Extending the prior to dynamic scenes might help in video-based correspondence.
  • The symmetry annotations could be used to test if the method respects object symmetries naturally.

Load-bearing premise

Disentangling canonical shape, deformation, and object pose from image reconstruction losses alone produces consistent semantic 3D correspondences across instances without any correspondence supervision or post-hoc alignment.

What would settle it

Training the model on HouseCorr3D and finding that predicted 3D keypoints fail to align with ground-truth annotations across different instances would falsify the emergence of correspondences.

Figures

Figures reproduced from arXiv: 2605.28257 by Adam Kortylewski, Artur Jesslen, Basavaraj Sunagad, Leonhard Sommer.

Figure 1
Figure 1. Figure 1: Monocular Category-level 3D Correspondence. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Dataset Overview. We annotate up to 19 3D keypoints directly on CAD meshes for 5–13 instances per category, covering 50 common household object classes. The keypoints are chosen to be semantically consistent and shared across all instances within each category We visualize a subset of these annotations across several categories to highlight their cross-instance and cross-shape consistency. Visualizations f… view at source ↗
Figure 3
Figure 3. Figure 3: (a) Monocular category-level 3D correspondence. Given a query point x q ∈ R 3 , we project it onto the deformed query mesh M q def and encode its location as barycentric coordinates. Since query and target instances share the same mesh topology, these coordinates transfer directly to Mt def , yielding the corresponding point x t ∈ R 3 . (b) Pipeline. Given an RGB-D image, the deformation encoder ψl predict… view at source ↗
Figure 4
Figure 4. Figure 4: Qualitative results. We compare 2D feature matching method DINOv2 [32], with 3D space matching methods GenPose++ [60], MagicPony [49], and Morpheus. For DINOv2 and GenPose++ we visualize the 2D correspondences. For MagicPony and Morpheus, we visualize the predicted deformed meshes in camera space, along with overlaid correspondence lines (see Sec. H). MagicPony’s predictions may appear visually plausible w… view at source ↗
read the original abstract

Understanding 3D objects from images is fundamental to robotics and AR/VR applications. While recent work has made progress in category-level pose estimation, current representations fail to capture the fine-grained semantics needed for reasoning about object parts, functions, and interactions. In this work, we study category-level 3D correspondence in camera space -- predicting, from a single image, 3D locations that remain consistent across instances within a category -- and show that it can emerge without explicit correspondence supervision by learning a shared morphable object prior. To enable research in this direction, we introduce HouseCorr3D, the first large-scale benchmark for monocular category-level 3D correspondence with 178k images across 50 household object categories, 280 unique instances, and 3D keypoint annotations directly on CAD models. Crucially, HouseCorr3D provides amodal correspondence labels for occluded regions and explicit symmetry annotations, addressing key limitations of existing datasets. We further propose Morpheus, a method that learns morphable category-level shape priors by disentangling canonical shape, deformation, and object pose. Through this shared canonical grounding, semantically meaningful 3D correspondences in camera space emerge implicitly. These emerging 3D correspondences set a new state of the art on HouseCorr3D, demonstrating that semantic 3D object understanding can arise without direct correspondence supervision. Data and code are publicly available at https://github.com/GenIntel/HouseCorr3D.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The paper claims that category-level 3D correspondences in camera space can emerge implicitly without explicit supervision by learning a shared morphable object prior (Morpheus) that disentangles canonical shape, deformation, and object pose from image reconstruction and pose objectives. To support evaluation, the authors introduce HouseCorr3D, a benchmark of 178k images across 50 household categories and 280 instances with 3D keypoint annotations on CAD models, including amodal labels for occlusions and explicit symmetry annotations. The method is reported to achieve state-of-the-art performance on this benchmark.

Significance. If the central claim holds, the work would be significant for demonstrating that semantic 3D object understanding can arise from standard reconstruction losses via a canonical representation, with implications for scalable learning in robotics and AR/VR. The benchmark's emphasis on amodal and symmetric cases addresses practical limitations of prior datasets. Public release of data and code is a clear strength that supports reproducibility.

major comments (3)
  1. [Abstract] Abstract: The claim that 'semantically meaningful 3D correspondences in camera space emerge implicitly' through disentangling canonical shape, deformation, and pose rests on the morphable prior discovering consistent semantics. However, pure image reconstruction losses are invariant to which semantic parts are assigned to which canonical coordinates, so nothing in the stated objectives prevents geometrically valid but semantically inconsistent alignments across instances (e.g., leg of one chair mapped to seat of another).
  2. [Method (Morpheus)] Method description: The disentanglement into canonical shape, deformation, and object pose is presented as producing the shared grounding for correspondences. No auxiliary semantic losses, part-aware features, or cross-instance regularizers are indicated in the abstract, making the emergence of semantics an empirical outcome of training rather than a property enforced by the formulation; ablations isolating the morphable prior's role are needed to substantiate the claim.
  3. [HouseCorr3D benchmark and evaluation] Benchmark and results: HouseCorr3D supplies 3D keypoint annotations directly on CAD models with amodal and symmetry handling. To support the 'without explicit correspondence supervision' advantage, the evaluation should include comparisons against supervised correspondence baselines and report cross-instance consistency metrics in addition to standard accuracy, as the predefined CAD keypoints may interact with the benchmark construction choices.
minor comments (2)
  1. [Abstract] Abstract: The SOTA claim would benefit from specifying the primary metric (e.g., PCK@threshold or mean distance error) on which Morpheus outperforms prior work.
  2. The distribution of the 178k images across the 280 instances and 50 categories should be detailed to allow assessment of category balance and instance diversity.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive comments. We address each major point below and will incorporate revisions to clarify the emergence claim, add ablations, and strengthen the evaluation.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The claim that 'semantically meaningful 3D correspondences in camera space emerge implicitly' through disentangling canonical shape, deformation, and pose rests on the morphable prior discovering consistent semantics. However, pure image reconstruction losses are invariant to which semantic parts are assigned to which canonical coordinates, so nothing in the stated objectives prevents geometrically valid but semantically inconsistent alignments across instances (e.g., leg of one chair mapped to seat of another).

    Authors: We agree that pure reconstruction losses are invariant to semantic labelings and do not explicitly prevent inconsistent part mappings. The emergence of consistent semantics is an empirical outcome driven by the requirement that a single shared canonical shape must explain all instances in the category. We will revise the abstract to state more precisely that correspondences emerge from the morphable prior under reconstruction and pose objectives, and add a short discussion paragraph explaining how the shared canonical representation promotes semantic consistency in practice despite the invariance. revision: yes

  2. Referee: [Method (Morpheus)] Method description: The disentanglement into canonical shape, deformation, and object pose is presented as producing the shared grounding for correspondences. No auxiliary semantic losses, part-aware features, or cross-instance regularizers are indicated in the abstract, making the emergence of semantics an empirical outcome of training rather than a property enforced by the formulation; ablations isolating the morphable prior's role are needed to substantiate the claim.

    Authors: The abstract is a high-level summary; the method section describes the architecture without auxiliary semantic losses. We agree that isolating ablations are required. In revision we will add experiments ablating the shared canonical shape (e.g., per-instance shape codes) and the disentanglement components, measuring impact on correspondence accuracy to substantiate the prior's role. revision: yes

  3. Referee: [HouseCorr3D benchmark and evaluation] Benchmark and results: HouseCorr3D supplies 3D keypoint annotations directly on CAD models with amodal and symmetry handling. To support the 'without explicit correspondence supervision' advantage, the evaluation should include comparisons against supervised correspondence baselines and report cross-instance consistency metrics in addition to standard accuracy, as the predefined CAD keypoints may interact with the benchmark construction choices.

    Authors: We will add supervised correspondence baselines trained on the HouseCorr3D keypoint annotations for direct comparison, while clearly noting that our method uses none of this supervision. We will also report cross-instance consistency (e.g., variance of predicted 3D locations for the same semantic keypoint across instances). We will clarify in the benchmark section that CAD models and keypoints were drawn from existing public sources independently of our method. revision: yes

Circularity Check

0 steps flagged

No significant circularity; central claim is empirical emergence from training

full rationale

The paper claims that category-level 3D correspondences emerge implicitly from disentangling canonical shape, deformation, and pose via image reconstruction losses in the Morpheus model, without explicit supervision. This is presented as a trained outcome on the new HouseCorr3D benchmark rather than a closed-form derivation. No equations, self-definitions, or fitted-input-as-prediction reductions are visible in the provided text. The result does not reduce by construction to its inputs; it is an observed property of the optimized network. No load-bearing self-citations or uniqueness theorems from prior author work are invoked in the abstract. This is the normal non-circular case for an empirical method paper.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 1 invented entities

The central claim rests on the empirical success of disentanglement in a neural network trained on category-level image data; the ledger records the modeling assumptions and the new prior construct.

free parameters (1)
  • latent dimensions for canonical shape and deformation
    Chosen hyperparameters that control the capacity of the morphable prior; their values are not derived from first principles.
axioms (1)
  • domain assumption A neural network trained on image reconstruction and pose objectives can learn a disentangled representation of canonical shape, deformation, and pose.
    Invoked when the abstract states that the shared canonical grounding produces correspondences.
invented entities (1)
  • Morphable object prior no independent evidence
    purpose: Provides the shared canonical reference that allows correspondences to emerge without explicit labels.
    New construct introduced by the method; no independent evidence outside the benchmark performance is supplied.

pith-pipeline@v0.9.1-grok · 5803 in / 1334 out tokens · 49224 ms · 2026-06-29T13:53:32.521689+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

64 extracted references · 17 canonical work pages · 1 internal anchor

  1. [1]

    In: Proceedings of the 26th Annual Conference on Computer Graphics and Inter- active Techniques

    Blanz, V., Vetter, T.: A morphable model for the synthesis of 3d faces. In: Proceedings of the 26th Annual Conference on Computer Graphics and Inter- active Techniques. p. 187–194. SIGGRAPH ’99, ACM Press/Addison-Wesley Publishing Co., USA (1999).https://doi.org/10.1145/311535.311556, https: //doi.org/10.1145/311535.3115564

    work page doi:10.1145/311535.311556 1999

  2. [2]

    In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

    Brazil, G., Kumar, A., Straub, J., Ravi, N., Johnson, J., Gkioxari, G.: Omni3D: A large benchmark and model for 3D object detection in the wild. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, Vancouver, Canada (June 2023) 4

    2023

  3. [3]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) (2021) 3

    Caron, M., Touvron, H., Misra, I., Jégou, H., Mairal, J., Bojanowski, P., Joulin, A.: Emerging properties in self-supervised vision transformers. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) (2021) 3

    2021

  4. [4]

    Chang, A.X., Funkhouser, T., Guibas, L., Hanrahan, P., Huang, Q., Li, Z., Savarese, S., Savva, M., Song, S., Su, H., Xiao, J., Yi, L., Yu, F.: ShapeNet: An Information- Rich 3D Model Repository. Tech. Rep. arXiv:1512.03012 [cs.GR], Stanford Univer- sity — Princeton University — Toyota Technological Institute at Chicago (2015) 4

    work page internal anchor Pith review Pith/arXiv arXiv 2015

  5. [5]

    In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

    Chen, X., Mottaghi, R., Liu, X., Fidler, S., Urtasun, R., Yuille, A.: Detect what you can: Detecting and representing objects using holistic models and body parts. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp. 1971–1978 (2014) 6

    1971

  6. [6]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) (2025) 4

    Dünkel, O., Wimmer, T., Theobalt, C., Rupprecht, C., Kortylewski, A.: Do it yourself: Learning semantic correspondence from pseudo-labels. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) (2025) 4

    2025

  7. [7]

    In: Conference on Neural Informa- tion Processing Systems (NeurIPS) (2022) 2

    Fu, Y., Wang, X.: Category-level 6d object pose estimation in the wild: A semi- supervised learning approach and a new dataset. In: Conference on Neural Informa- tion Processing Systems (NeurIPS) (2022) 2

    2022

  8. [8]

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

    Gkioxari, G., Johnson, J., Malik, J.: Mesh r-cnn. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). pp. 9784–9794 (2019).https: //doi.org/10.1109/ICCV.2019.009884

    work page doi:10.1109/iccv.2019.009884 2019

  9. [9]

    In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

    Goel, S., Gkioxari, G., Malik, J.: Differentiable stereopsis: Meshes from multiple views using differentiable rendering. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp. 8635–8644 (2022) 9

    2022

  10. [10]

    In: Proceedings of the International Conference on Machine Learning (ICML) (2020) 10, 29 16 Sommer et al

    Gropp, A., Yariv, L., Haim, N., Atzmon, M., Lipman, Y.: Implicit geometric regularization for learning shapes. In: Proceedings of the International Conference on Machine Learning (ICML) (2020) 10, 29 16 Sommer et al

    2020

  11. [11]

    In: European Conference on Computer Vision (ECCV) (2018) 4

    Groueix, T., Fisher, M., Kim, V.G., Russell, B., Aubry, M.: 3d-coded : 3d corre- spondences by deep deformation. In: European Conference on Computer Vision (ECCV) (2018) 4

    2018

  12. [12]

    In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

    Güler, R.A., Neverova, N., Kokkinos, I.: Densepose: Dense human pose estimation in the wild. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp. 7297–7306 (2018) 4

    2018

  13. [13]

    In: IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI) (2016) 3, 5, 6

    Ham, B., Cho, M., Schmid, C., Ponce, J.: Proposal flow: Semantic correspondences from object proposals. In: IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI) (2016) 3, 5, 6

    2016

  14. [14]

    In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

    He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp. 770–778 (2016).https://doi.org/10.1109/CVPR.2016.9011, 23

    work page doi:10.1109/cvpr.2016.9011 2016

  15. [15]

    Reiss, N

    Jakab, T., Tucker, R., Makadia, A., Wu, J., Snavely, N., Kanazawa, A.: Keypoint- deformer: Unsupervised 3d keypoint discovery for shape control. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp. 12778–12787 (2021).https://doi.org/10.1109/CVPR46437.2021.012594

    work page doi:10.1109/cvpr46437.2021.012594 2021

  16. [16]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) (2021) 2, 4

    Jiang, W., Trulls, E., Hosang, J., Tagliasacchi, A., Yi, K.M.: COTR: Correspon- dence Transformer for Matching Across Images. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) (2021) 2, 4

    2021

  17. [17]

    In: International Conference on 3D Vision (3DV) (2025) 4, 30

    Kim, H., Lang, I., Aigerman, N., Groueix, T., Kim, V.G., Hanocka, R.: Meshup: Multi-target mesh deformation via blended score distillation. In: International Conference on 3D Vision (3DV) (2025) 4, 30

    2025

  18. [18]

    In: International Conference on Learning Representations (ICLR) (2015),https://arxiv.org/abs/ 1412.698011

    Kingma, D.P., Ba, J.: Adam: A method for stochastic optimization. In: International Conference on Learning Representations (ICLR) (2015),https://arxiv.org/abs/ 1412.698011

    work page arXiv 2015

  19. [19]

    In: European Conference on Computer Vision (ECCV) (2024) 2, 5

    Krishnan, A., Kundu, A., Maninis, K.K., Hays, J., Brown, M.: Omninocs: A unified nocs dataset and model for 3d lifting of 2d objects. In: European Conference on Computer Vision (ECCV) (2024) 2, 5

    2024

  20. [20]

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

    Kulkarni, N., Tulsiani, S., Gupta, A.: Canonical surface mapping via geometric cycle consistency. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). pp. 2202–2211 (2019).https://doi.org/10.1109/ICCV. 2019.002294

    work page doi:10.1109/iccv 2019

  21. [21]

    arXiv preprint arXiv:2305.02385 (2023) 3

    Li, X., Han, K., Wan, X., Prisacariu, V.A.: Simsc: A simple framework for semantic correspondence with temperature learning. arXiv preprint arXiv:2305.02385 (2023) 3

    work page arXiv 2023

  22. [22]

    IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI)33(5), 978–994 (2011) 3

    Liu, C., Yuen, J., Torralba, A.: Sift flow: Dense correspondence across scenes and its applications. IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI)33(5), 978–994 (2011) 3

    2011

  23. [23]

    ACM Trans

    Loper, M., Mahmood, N., Romero, J., Pons-Moll, G., Black, M.J.: SMPL: A skinned multi-person linear model. ACM Trans. Graphics (Proc. SIGGRAPH Asia)34(6), 248:1–248:16 (Oct 2015) 4

    2015

  24. [24]

    In: European Conference on Computer Vision (ECCV)

    Lou, Y., You, Y., Li, C., Cheng, Z., Li, L., Ma, L., Wang, W., Lu, C.: Human correspondence consensus for 3d object semantic understanding. In: European Conference on Computer Vision (ECCV). p. 496–512. Springer-Verlag, Berlin, Heidelberg (2020). https://doi.org/10.1007/978-3-030-58542-6_30 , https: //doi.org/10.1007/978-3-030-58542-6_304, 6

    work page doi:10.1007/978-3-030-58542-6_30 2020

  25. [25]

    International Journal of Computer Vision (IJCV)60(2), 91–110 (2004) 3

    Lowe, D.G.: Distinctive image features from scale-invariant keypoints. International Journal of Computer Vision (IJCV)60(2), 91–110 (2004) 3

    2004

  26. [26]

    In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

    Mariotti, O., Mac Aodha, O., Bilen, H.: Improving semantic correspondence with viewpoint-guided spherical maps. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp. 19521–19530 (2024) 2, 4 HouseCorr3D: Category-Level 3D Correspondence 17

    2024

  27. [27]

    Min, J., Lee, J., Ponce, J., Cho, M.: Spair-71k: A large-scale benchmark for semantic correspondence (2019),https://arxiv.org/abs/1908.105432, 3, 5, 6, 13

    work page arXiv 2019

  28. [28]

    In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2019) 4

    Mo, K., Zhu, S., Chang, A.X., Yi, L., Tripathi, S., Guibas, L., Su, H.: Partnet: A large-scale benchmark for fine-grained and hierarchical part-level 3d object understanding. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2019) 4

    2019

  29. [29]

    In: International Conference on Learning Representations (ICLR)

    Nam, J., Lee, G., Kim, S., Kim, H., Cho, H., Kim, S., Kim, S.: Diffusion model for dense matching. In: International Conference on Learning Representations (ICLR). OpenReview.net (2024),https://openreview.net/forum?id=Zsfiqpft6K2, 4

    2024

  30. [30]

    Conference on Neural Information Processing Systems (NeurIPS) (2020) 4, 30

    Neverova, N., Novotny, D., Khalidov, V., Szafraniec, M., Labatut, P., Vedaldi, A.: Continuous surface embeddings for deformable shape correspondence. Conference on Neural Information Processing Systems (NeurIPS) (2020) 4, 30

    2020

  31. [31]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) (2019) 4

    Novotny, D., Ravi, N., Graham, B., Neverova, N., Vedaldi, A.: C3dpo: Canonical 3d pose networks for non-rigid structure from motion. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) (2019) 4

    2019

  32. [32]

    Oquab, M., Darcet, T., Moutakanni, T., Vo, H.V., Szafraniec, M., Khalidov, V., Fernandez, P., Haziza, D., Massa, F., El-Nouby, A., Howes, R., Huang, P.Y., Xu, H., Sharma, V., Li, S.W., Galuba, W., Rabbat, M., Assran, M., Ballas, N., Synnaeve, G., Misra, I., Jegou, H., Mairal, J., Labatut, P., Joulin, A., Bojanowski, P.: Dinov2: Learning robust visual feat...

    2023

  33. [33]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) (2021) 4

    Reizenstein, J., Shapovalov, R., Henzler, P., Sbordone, L., Labatut, P., Novotny, D.: Common objects in 3d: Large-scale learning and evaluation of real-life 3d category reconstruction. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) (2021) 4

    2021

  34. [34]

    Conference on Neural Information Processing Systems (NeurIPS)34, 6087–6101 (2021) 9

    Shen, T., Gao, J., Yin, K., Liu, M.Y., Fidler, S.: Deep marching tetrahedra: a hybrid representation for high-resolution 3d shape synthesis. Conference on Neural Information Processing Systems (NeurIPS)34, 6087–6101 (2021) 9

    2021

  35. [35]

    In: European Conference on Computer Vision (ECCV) (2024) 4, 30, 31

    Shtedritski, A., Rupprecht, C., Vedaldi, A.: Shic: Shape-image correspondences with no keypoint supervision. In: European Conference on Computer Vision (ECCV) (2024) 4, 30, 31

    2024

  36. [36]

    In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

    Sommer, L., Dünkel, O., Theobalt, C., Kortylewski, A.: Common3d: Self-supervised learning of 3d morphable models for common objects in neural feature space. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp. 6468–6479 (June 2025) 4, 22

    2025

  37. [37]

    Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2021) 2, 4

    Sun, J., Shen, Z., Wang, Y., Bao, H., Zhou, X.: LoFTR: Detector-free local feature matching with transformers. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2021) 2, 4

    2021

  38. [38]

    In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2018) 4

    Sun, X., Wu, J., Zhang, X., Zhang, Z., Zhang, C., Xue, T., Tenenbaum, J.B., Freeman, W.T.: Pix3d: Dataset and methods for single-image 3d shape modeling. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2018) 4

    2018

  39. [39]

    In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

    Sun, Y., Huang, Y., Guo, H., Zhao, Y., Wu, R., Yu, Y., Ge, W., Zhang, W.: Misc210k: A large-scale dataset for multi-instance semantic correspondence. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp. 7121–7130 (2023) 5, 6

    2023

  40. [40]

    In: Bengio, S., Wallach, H.M., Larochelle, H.,Grauman,K.,Cesa-Bianchi,N.,Garnett,R.(eds.)ConferenceonNeuralInforma- tion Processing Systems (NeurIPS)

    Suwajanakorn, S., Snavely, N., Tompson, J., Norouzi, M.: Discovery of latent 3d key- points via end-to-end geometric reasoning. In: Bengio, S., Wallach, H.M., Larochelle, H.,Grauman,K.,Cesa-Bianchi,N.,Garnett,R.(eds.)ConferenceonNeuralInforma- tion Processing Systems (NeurIPS). pp. 2063–2074 (2018),https://proceedings. neurips.cc/paper/2018/hash/24146db4e...

    2063

  41. [41]

    In: European Conference on Computer Vision (ECCV)

    Tian, M., Ang, M.H., Lee, G.H.: Shape prior deformation for categorical 6d object pose and size estimation. In: European Conference on Computer Vision (ECCV). p. 530–546. Springer-Verlag, Berlin, Heidelberg (2020).https://doi.org/10.1007/ 978-3-030-58589-1_32,https://doi.org/10.1007/978-3-030-58589-1_324

    work page doi:10.1007/978-3-030-58589-1_324 2020

  42. [42]

    IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI)32, 815–30 (05 2010).https://doi.org/10.1109/TPAMI.2009.773

    Tola, E., Lepetit, V., Fua, P.: Daisy: An efficient dense descriptor applied to wide baseline stereo. IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI)32, 815–30 (05 2010).https://doi.org/10.1109/TPAMI.2009.773

    work page doi:10.1109/tpami.2009.773 2010

  43. [43]

    Wah, C., Branson, S., Welinder, P., Perona, P., Belongie, S.: Cub-200-2011 (Apr 2022).https://doi.org/10.22002/D1.200984, 5, 6

    work page doi:10.22002/d1.200984 2011

  44. [44]

    Showui: One vision-language- action model for GUI visual agent

    Wandel, K., Wang, H.: Semalign3d: Semantic correspondence between rgb-images through aligning 3d object-class representations. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp. 1138–1147 (2025).https://doi.org/10.1109/CVPR52734.2025.001144

    work page doi:10.1109/cvpr52734.2025.001144 2025

  45. [45]

    In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

    Wang, H., Sridhar, S., Huang, J., Valentin, J., Song, S., Guibas, L.J.: Normalized object coordinate space for category-level 6d object pose and size estimation. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp. 2642–2651 (2019) 2, 5, 11, 12, 21, 23, 30

    2019

  46. [46]

    In: European Conference on Computer Vision (ECCV) (2018) 4

    Wang, N., Zhang, Y., Li, Z., Fu, Y., Liu, W., Jiang, Y.G.: Pixel2mesh: Generating 3d mesh models from single rgb images. In: European Conference on Computer Vision (ECCV) (2018) 4

    2018

  47. [47]

    In: Proceedings of the IEEE/CVF Interna- tional Conference on Computer Vision (ICCV)

    Weinzaepfel, P., Revaud, J., Harchaoui, Z., Schmid, C.: Deepflow: Large displace- ment optical flow with deep matching. In: Proceedings of the IEEE/CVF Interna- tional Conference on Computer Vision (ICCV). pp. 1385–1392 (2013) 3

    2013

  48. [48]

    International Journal of Computer Vision (IJCV) (2023) 4, 5, 6

    Wu, S., Jakab, T., Rupprecht, C., Vedaldi, A.: DOVE: Learning deformable 3d objects by watching videos. International Journal of Computer Vision (IJCV) (2023) 4, 5, 6

    2023

  49. [49]

    In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2023) 4, 9, 10, 11, 12, 22, 23

    Wu, S., Li, R., Jakab, T., Rupprecht, C., Vedaldi, A.: MagicPony: Learning articu- lated 3d animals in the wild. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2023) 4, 9, 10, 11, 12, 22, 23

    2023

  50. [50]

    In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2015) 4

    Wu, Z., Song, S., Khosla, A., Yu, F., Zhang, L., Tang, X., Xiao, J.: 3d shapenets: A deep representation for volumetric shapes. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2015) 4

    2015

  51. [51]

    In: Proceedings of the Winter Conference on Applications of Computer Vision (WACV)

    Xiang, Y., Mottaghi, R., Savarese, S.: Beyond pascal: A benchmark for 3d object detection in the wild. In: Proceedings of the Winter Conference on Applications of Computer Vision (WACV). pp. 75–82. IEEE (2014) 4

    2014

  52. [52]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) (2023) 4

    Xu, J., Zhang, Y., Peng, J., Ma, W., Jesslen, A., Ji, P., Hu, Q., Zhang, J., Liu, Q., Wang, J., Ji, W., Wang, C., Yuan, X., Kaushik, P., Zhang, G., Liu, J., Xie, Y., Cui, Y., Yuille, A., Kortylewski, A.: Animal3d: A comprehensive dataset of 3d animal pose and shape. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) (2023) 4

    2023

  53. [53]

    In: Conference on Robot Learning (CoRL) (2020) 3

    Xu, Z., He, Z., Wu, J., Song, S.: Learning 3d dynamic scene representations for robot manipulation. In: Conference on Robot Learning (CoRL) (2020) 3

    2020

  54. [54]

    In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

    Yang, G., Sun, D., Jampani, V., Vlasic, D., Cole, F., Chang, H., Ramanan, D., Freeman, W.T., Liu, C.: Lasr: Learning articulated shape reconstruction from a monocular video. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp. 15980–15989 (2021) 9

    2021

  55. [55]

    In: ACM Trans

    Yi, L., Kim, V.G., Ceylan, D., Shen, W., Yan, M., Su, H., Lu, C., Huang, Q., Sheffer, A., Guibas, L.: A scalable active framework for region annotation in 3d shape collections. In: ACM Trans. Graphics (Proc. SIGGRAPH Asia) (2016) 4 HouseCorr3D: Category-Level 3D Correspondence 19

    2016

  56. [56]

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

    Yifan, W., Aigerman, N., Kim, V.G., Chaudhuri, S., Sorkine-Hornung, O.: Neural cages for detail-preserving 3d deformations. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp. 72–80 (2020).https: //doi.org/10.1109/CVPR42600.2020.000154

    work page doi:10.1109/cvpr42600.2020.000154 2020

  57. [57]

    Nature Human Behaviour8(2), 320–335 (Feb 2024) 3

    Yildirim, I., Siegel, M.H., Soltani, A.A., Ray Chaudhuri, S., Tenenbaum, J.B.: Perception of 3D shape integrates intuitive physics and analysis-by-synthesis. Nature Human Behaviour8(2), 320–335 (Feb 2024) 3

    2024

  58. [58]

    IEEE Trans- actions on Pattern Analysis and Machine Intelligence (PAMI)44(9), 5780–5795 (2022).https://doi.org/10.1109/TPAMI.2021.30726594

    You, Y., Li, C., Lou, Y., Cheng, Z., Li, L., Ma, L., Wang, W., Lu, C.: Understanding pixel-level 2d image semantics with 3d keypoint knowledge engine. IEEE Trans- actions on Pattern Analysis and Machine Intelligence (PAMI)44(9), 5780–5795 (2022).https://doi.org/10.1109/TPAMI.2021.30726594

    work page doi:10.1109/tpami.2021.30726594 2022

  59. [59]

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

    You, Y., Lou, Y., Li, C., Cheng, Z., Li, L., Ma, L., Lu, C., Wang, W.: Keypointnet: A large-scale 3d keypoint dataset aggregated from numerous human annotations. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp. 13644–13653 (2020).https://doi.org/10.1109/CVPR42600.2020. 013664, 6

    work page doi:10.1109/cvpr42600.2020 2020

  60. [60]

    In: European Conference on Computer Vision (ECCV)

    Zhang, J., Huang, W., Peng, B., Wu, M., Hu, F., Chen, Z., Zhao, B., Dong, H.: Omni6dpose: Large-scale multi-object 6d pose estimation with realistic rendering. In: European Conference on Computer Vision (ECCV). pp. 3110–3120 (2024) 2, 4, 6, 7, 8, 11, 12, 22, 23

    2024

  61. [61]

    In: Conference on Neural Information Processing Systems (NeurIPS) (2023) 3, 23

    Zhang, J., Herrmann, C., Hur, J., Cabrera, L.P., Jampani, V., Sun, D., Yang, M.H.: A Tale of Two Features: Stable Diffusion Complements DINO for Zero- Shot Semantic Correspondence. In: Conference on Neural Information Processing Systems (NeurIPS) (2023) 3, 23

    2023

  62. [62]

    In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

    Zheng, Z., Yu, T., Dai, Q., Liu, Y.: Deep implicit templates for 3d shape represen- tation. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp. 1429–1439 (2021) 9, 10, 30

    2021

  63. [63]

    International Conference on Learning Representations (ICLR) (2022) 3

    Zhou, J., Wei, C., Wang, H., Shen, W., Xie, C., Yuille, A., Kong, T.: ibot: Image bert pre-training with online tokenizer. International Conference on Learning Representations (ICLR) (2022) 3

    2022

  64. [64]

    International Conference on Learning Representations (ICLR) (2025) 4, 5, 6, 30 20 Sommer et al

    Zhu, J., Ju, Y., Zhang, J., Wang, M., Yuan, Z., Hu, K., Xu, H.: Densematcher: Learning 3d semantic correspondence for category-level manipulation from a single demo. International Conference on Learning Representations (ICLR) (2025) 4, 5, 6, 30 20 Sommer et al. Category-Level 3D Correspondence in Camera Space via Morphable Object Priors Supplementary Mate...

    2025