CorNet: Generic 3D Corners for 6D Pose Estimation of New Objects without Retraining

Giorgia Pitteri; Slobodan Ilic; Vincent Lepetit

REVIEW 4 major objections 5 minor 39 references

CorNet: Generic 3D Corners for 6D Pose Estimation of New Objects without Retraining

T0 review · 4 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read A corner detector trained once on a handful of objects can locate and pose unseen industrial objects from their CAD models alone, without retraining.

desk verdict Promising idea for object-agnostic 6D pose from RGB, but the evidence is too thin and the corner-ambiguity assumption needs explicit justification. read the letter →

arxiv 1908.11457 v1 pith:YMXKKUPO submitted 2019-08-29 cs.CV

Giorgia Pitteri , Slobodan Ilic , Vincent Lepetit This is my paper · ORCID

classification cs.CV

keywords 6Dposeestimationgenericcornerdetectionobject-agnosticCAD-model-basedRANSACmatchingambiguityT-LESSdatasetindustrialobjects

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

The paper tries to establish that a single offline-trained corner detector is enough to detect and estimate the 6D pose of objects never seen during training, as long as their CAD models are available. This matters because industrial pose-estimation systems normally require per-object retraining on many registered images; replacing that with a CAD file removes a major bottleneck. The method treats corners as generic reusable parts: it detects them in the image, predicts their 3D poses from seven virtual control points, then uses a RANSAC-like matcher to align those corners with corners extracted from the new object's CAD model. The paper demonstrates the idea on the T-LESS dataset with a train/test split where the test objects are never shown to the network.

What carries the argument

The central object is the generic 3D corner represented by seven virtual control points arranged around the corner and spanning three orthogonal directions. The pose of a corner is predicted as the 2D reprojections of these points, which makes it easy to combine corner hypotheses through a PnP solve. The two permutations Sigma1 and Sigma2, which rearrange the reprojected points, enumerate the three possible 3D poses that are visually consistent with a given corner image, and this hypothesis set is fed into the RANSAC-like matcher along with gradient-based scoring to select the final object pose.

What would settle it

Render a single corner from all viewpoints and cluster the renders by appearance; if more than three distinct 3D poses share the same appearance for some corner geometry, the two-permutation hypothesis set is incomplete. Feeding that corner through the full pipeline and comparing against a brute-force search over all corner rotations would then show the method missing valid poses.

Watch

Extended reading notes

Core claim

Given only the CAD model of a new object with corners, the method detects that object in a single color image and estimates its 6D pose without any new training phase. A Faster R-CNN-style detector is trained once, on synthetic images of a small set of objects, to find generic corners and predict each corner's 3D pose in the form of 2D reprojections of seven virtual control points. Because a corner's appearance can correspond to several 3D poses, the paper introduces two permutations of the reprojected points to generate the complete set of pose hypotheses; a RANSAC-like procedure then matches these hypotheses to the corners extracted from the CAD model, computes a pose via PnP, refines using inliers, and scores the result by cross-correlating image gradients with the rendered CAD model. Only one or two detected corners are needed to estimate the object pose, which gives the method resilience to partial occlusion.

Load-bearing premise

The argument assumes that every detectable corner has exactly three possible 3D poses that look the same in an image, generated by the two permutations; if a corner's symmetry yields more or different ambiguities, the RANSAC matcher cannot propose the correct pose.

Editorial extensions

If this is right

A new industrial object can be added to a pose-estimation system by supplying its CAD model; no training images, domain adaptation, or network retraining are needed.
Because one or two detected corners suffice for a PnP pose estimate, the pipeline remains usable under partial occlusion and clutter, as the qualitative T-LESS results illustrate.
The same trained corner vocabulary transfers across different object instances, at least for objects whose corners resemble those seen during training.
When multiple CAD objects share the same corner arrangement, the final full-3D-geometry check decides which object and pose is present.

Reading between the lines

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

Editorial inference: if the generic corner vocabulary is the active ingredient, the same training set should transfer to other industrial object families beyond T-LESS; this is testable by training once on one set of objects and evaluating on CAD models from a different dataset.
Editorial inference: the exactly-three-poses model is a design choice, so objects with higher-symmetry corners (for example, fourfold or regular polyhedral corners) would plausibly need additional permutations to keep the RANSAC hypothesis set complete.
Editorial inference: the method suggests a parts-composition route to pose estimation in which learned reusable parts plus geometric matching replace full-object descriptors; the authors mention edges and quadric surfaces as natural next parts, which would broaden coverage beyond corner-rich objects.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Colleague,

Here's my read on the CorNet paper. The thing to know: it does something genuinely new—learning generic corner detectors and matching them to CAD models to estimate 6D pose of objects never seen in training—but the experimental case is much weaker than the title suggests. The average ADD(10%) is 34.7% on five T-LESS objects with no baselines, no ablations, and no code. I'd treat it as a promising early-stage idea, not a validated method.

The core idea is real. Using the virtual-control-point representation from Crivellaro et al. and extending it to generic corners with explicit handling of symmetry ambiguities is a sensible step. The RANSAC-like matching that needs only one or two corners is a nice practical contribution, and the final gradient-based verification is a reasonable way to choose among hypotheses. The paper is clearly written, and the reliance on the authors' own prior part representation is legitimate; it's a building block, not a reused result.

The soft spots are where I'd push. Section 3.2 asserts that a corner has exactly three poses and generates only {I, Sigma1, Sigma2} hypotheses. For a cube corner that's right: the 120-degree rotations around the body diagonal. For a generic industrial corner—unequal edge lengths, chamfers, more than three faces—the symmetry group can be trivial or something else, and the projection-to-pose ambiguity isn't necessarily three. If the hypothesis set is incomplete, the matcher can never find the true pose. The paper doesn't justify why three is enough for the corners it actually uses. Relatedly, the ordering of the seven control points between the network's output and the CAD-model corner has to be consistent; a transposition of two axes is a reflection, not a rotation, and would be missed by the two permutations. The authors probably used a fixed convention, but they don't state it. These are fixable with clarification and possibly a small expansion of the hypothesis set, but they're load-bearing enough that they need explicit treatment.

The evaluation is the bigger problem. One split, five test objects, no comparison to any alternative, no ablation of the corner detector or the matching, no code, and the inlier threshold is never specified. For a claim this strong, that's not enough.

Who's it for? Someone working in RGB 6D pose estimation, especially in robotics or industrial vision, will want to read this; the problem is practical and the approach is a real departure from per-object training. It deserves a serious referee, but I'd send it back with clear demands for baselines, ablations, code, and a real discussion of the ambiguity assumption. Not a desk reject, not an accept.

Referee Report

4 major / 5 minor

Summary. The paper proposes CorNet, a method for 6D pose estimation of new, unseen objects from a single RGB image without retraining the network for those objects. The method trains a Faster R-CNN detector, once, on synthetic images of a small set of objects, to detect generic corners and predict their 3D poses in the form of 2D projections of seven virtual control points. At test time, the CAD model of a new object is used to define its 3D corners, and a RANSAC-like algorithm matches detected 2D corners to these 3D corners, generating three pose hypotheses per corner via two permutations Sigma1 and Sigma2 to account for corner symmetry ambiguities. A final gradient cross-correlation score selects the best pose. The method is evaluated on five unseen objects from the T-LESS dataset across eleven test scenes, reporting an average ADD(10%) accuracy of 34.7% and a detection accuracy of 52.5%.

Significance. If the central claim holds, the paper addresses a practically important and relatively unexplored problem: 6D pose estimation of objects never seen during training, using only their CAD geometry. This would be a useful step for industrial applications where re-training for each new part is costly. The paper has several strengths: it introduces a generic corner formulation based on the virtual control point representation, explicitly addresses corner pose ambiguity with a RANSAC-like algorithm, and reports quantitative results on the challenging T-LESS dataset. The main caveat is that the quantitative evidence is narrow and the ambiguity model is not validated; the contribution is therefore promising but not yet established at the level of a definitive claim.

major comments (4)

[Section 3.2 and Algorithm 1] The assertion that a corner image has 'in general 3 possible 3D poses' is used to restrict the hypothesis set to {I, Sigma1, Sigma2}. This is load-bearing because any valid pose outside this set can never be recovered by the subsequent PnP and scoring stages. The paper does not justify that this set is complete for the corner class considered: a general trihedral corner need not have a 3-fold symmetry, and corners with chamfers or additional incident faces can have different symmetry groups. Please either restrict the claim to the symmetry class actually present in the training and test corners, prove completeness for that class, and add experiments on corners whose symmetry group is not of order 3.
[Section 3.3, Pose Estimation Algorithm] The detected corner representation is an ordered list of seven 2D points, and the CAD corner is an ordered list of seven 3D virtual points, but the manuscript does not specify how the ordering is fixed for a new object's CAD model. The network learns a fixed labeling convention from the training corners; if the CAD ordering for a new corner differs by a transposition of two axes, the correspondence is wrong and the correct pose is not among the hypotheses generated by Sigma1 and Sigma2, because those two permutations generate only a cyclic subgroup. State the canonical ordering used for every new CAD corner and verify (for example, by checking invariance or by explicitly testing all six permutations) that the chosen ordering is compatible with the network's convention.
[Section 4.2, Table 1] The evaluation consists of a single hand-picked train/test split of five objects from T-LESS, with no baseline or ablation. The average ADD(10%) of 34.7% and detection accuracy of 52.5% are hard to interpret without comparison to a simple alternative, such as template matching on CAD renderings, a version of the pipeline that ignores the corner ambiguity, or a version that uses all six edge permutations. The paper's central claim is that the corner representation plus the ambiguity-handling algorithm enables pose estimation of unseen objects; an ablation study is needed to show that the reported results are not mostly due to the gradient-scoring stage alone. I also recommend reporting results over more than one split, or at least justifying why the chosen objects are representative.
[Section 4.2.2 and Table 1] The text says 'For symmetrical objects, we report the ADI metric instead of ADD', but Table 1 does not state which objects are treated as symmetric, and the metric name is still written as ADD in the column header. ADD and ADI are not interchangeable, and mixing them without a per-object annotation makes the pose accuracy numbers impossible to interpret or reproduce. Please mark per object which metric was used and define the exact ADI variant and threshold.

minor comments (5)

[Abstract] The abstract says 'does not require any training phases nor data for new objects'; this is clear in context, but consider rewording to 'no training phase for the new object' because the method does require an offline training phase on a small set of objects.
[Section 3.2, Figure 5] The text says 'permuting properly the 2D reprojections' but does not specify how Sigma1 and Sigma2 act on the seven control points; give the explicit permutations, for example as cycles over the control-point indices.
[Algorithm 1] The pseudocode uses the symbol '⊿' in a nonstandard way; use a standard comment delimiter and clarify the control flow for collecting multiple pose candidates before the final scoring.
[Table 1] The table reports standard deviations (for example, ±18.5) but it is not stated whether these are computed per scene, per object, or per frame; specify the unit of aggregation.
[General] The manuscript contains no statement about code or data availability; sharing the trained detector and the exact corner-extraction routine would be valuable for reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper evaluates pose estimation on CAD-only unseen objects, and the cited pose representation is an external building block.

full rationale

The derivation chain is self-contained: a corner detector and pose predictor are trained on a small set of T-LESS objects (#6, #19, #25, #27, #28), then evaluated on novel objects (#7, #8, #20, #26, #29) using only their CAD models. No test-object images or poses are used to fit the network, so the reported poses are genuine predictions rather than refitted values. The 7-point virtual-control-point representation is adopted from [7], which includes a coauthor of the present paper, but [7] is a published, independently evaluated representation used only as a building block; it does not by itself certify the no-retraining claim. The Section 3.2 assumption that a corner image yields exactly three pose hypotheses via permutations Sigma1 and Sigma2 is a modeling assumption that could be incomplete for some corner geometries, but it is not circular: the hypothesis set is not defined in terms of the target object pose, and the final gradient-based scoring is an independent comparison against the CAD rendering. Thus the central contribution is not equivalent to its inputs by construction.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

The method's central claim rests on a small number of domain assumptions, primarily that unseen objects have corner geometries and appearances similar to the training set, and that the three-fold pose ambiguity model is complete for those corners. The only fitted model parameters are the network weights, which are not released, and the RANSAC threshold is unreported.

free parameters (3)

Trained Faster R-CNN weights = Learned on synthetic images of T-LESS Objects #6, #19, #25, #27, #28
The central method depends on these learned detector and pose regression weights; the paper does not release them, so the fitted model itself is not independently reproducible.
Inlier threshold tau_inliers = not reported
Algorithm 1 uses this threshold to decide which corner matches are kept before refinement; its value affects pose accuracy and is not specified.
Number of control points Nv = 7
Chosen as in [7]; this design parameter determines the PnP conditioning and the size of the pose hypotheses.

assumptions (6)

domain assumption Objects of interest have prominent corners that can be extracted from the CAD model.
The method is scoped to corner-bearing objects in Section 1, and the evaluation only includes such objects.
domain assumption The appearance of corners from unseen objects is sufficiently similar to corners seen during training.
The entire transfer claim relies on Faster R-CNN generalizing across object instances; Section 3.1 states this empirically but does not quantify it.
ad hoc to paper A corner's 3D pose under perspective projection is ambiguous only up to the three rotations generated by Sigma1 and Sigma2.
Section 3.2 and Fig. 5 assert this for the corner shapes considered; other symmetry groups would require more hypotheses.
standard math PnP from seven 3D-2D correspondences yields a valid object pose candidate.
Used in Algorithm 1; this is a standard result in geometric computer vision [13].
ad hoc to paper The gradient cross-correlation score reliably ranks pose hypotheses.
Used as the final pose selection in the SCORE procedure; no ablation or analysis supports this choice.
domain assumption Synthetic renderings need no domain transfer to match real T-LESS images.
Section 3.1 states this was noticed in practice; it may not hold for other domains or image conditions.

how reviews work

0 comments

Cite this review

Pith. "Pith review of CorNet: Generic 3D Corners for 6D Pose Estimation of New Objects without Retraining." pith.science (2026). https://pith.science/paper/YMXKKUPO

@misc{pith2026190811457,
  author       = {Pith},
  title        = {Pith review of: CorNet: Generic 3D Corners for 6D Pose Estimation of New Objects without Retraining},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YMXKKUPO}},
  note         = {Machine review of arXiv:1908.11457}
}

read the original abstract

We present a novel approach to the detection and 3D pose estimation of objects in color images. Its main contribution is that it does not require any training phases nor data for new objects, while state-of-the-art methods typically require hours of training time and hundreds of training registered images. Instead, our method relies only on the objects' geometries. Our method focuses on objects with prominent corners, which covers a large number of industrial objects. We first learn to detect object corners of various shapes in images and also to predict their 3D poses, by using training images of a small set of objects. To detect a new object in a given image, we first identify its corners from its CAD model; we also detect the corners visible in the image and predict their 3D poses. We then introduce a RANSAC-like algorithm that robustly and efficiently detects and estimates the object's 3D pose by matching its corners on the CAD model with their detected counterparts in the image. Because we also estimate the 3D poses of the corners in the image, detecting only 1 or 2 corners is sufficient to estimate the pose of the object, which makes the approach robust to occlusions. We finally rely on a final check that exploits the full 3D geometry of the objects, in case multiple objects have the same corner spatial arrangement. The advantages of our approach make it particularly attractive for industrial contexts, and we demonstrate our approach on the challenging T-LESS dataset.

Figures

Figures reproduced from arXiv: 1908.11457 by the authors.

**Figure 2.** Overview of our approach. We modified Faster [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** 3D pose representation of an object part from [7]. [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗

Figures from the paper (8 more)

**Figure 5.** Figure 5: Given the image of a corner, three arrangements [PITH_FULL_IMAGE:figures/full_fig_p004_5.png]

**Figure 1.** Figure 1: These images are created by randomly placing the [PITH_FULL_IMAGE:figures/full_fig_p004_1.png]

**Figure 6.** Figure 6: Some qualitative results on Object #8 in Scene #03 of the T-LESS dataset. First row: 2D detection results. Second [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]

**Figure 7.** Figure 7: Some qualitative results on Object #7 in Scene #06 of the T-LESS dataset. [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]

**Figure 8.** Figure 8: Some qualitative results on Object #20 in Scene #10 of the T-LESS dataset. [PITH_FULL_IMAGE:figures/full_fig_p007_8.png]

**Figure 9.** Figure 9: Some qualitative results on Object #20 in Scene #13 of the T-LESS dataset. [PITH_FULL_IMAGE:figures/full_fig_p008_9.png]

**Figure 10.** Figure 10: Some qualitative results on Object #20 in Scene #14 of the T-LESS dataset. [PITH_FULL_IMAGE:figures/full_fig_p008_10.png]

**Figure 11.** Figure 11: Some qualitative results on Object #26 and Object #29 in Scene #15 of the T-LESS dataset. [PITH_FULL_IMAGE:figures/full_fig_p008_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 38 canonical work pages

[1]

Balntas, A

V . Balntas, A. Doumanoglou, C. Sahin, J. Sock, R. Kousk- ouridas, and T.-K. Kim. Pose Guided RGBD Feature Learn- ing for 3D Object Pose Estimation. In International Confer- ence on Computer Vision, 2017

work page 2017
[2]

Bousmalis, N

K. Bousmalis, N. Silberman, D. Dohan, D. Erhan, and D. Kr- ishnan. Unsupervised Pixel-Level Domain Adaptation with Generative Adversarial Networks. In Conference on Com- puter Vision and Pattern Recognition, 2017

work page 2017
[3]

Bousmalis, G

K. Bousmalis, G. Trigeorgis, N. Silberman, D. Krishnan, and D. Erhan. Domain Separation Networks. In Advances in Neural Information Processing Systems , pages 343–351, 2016

work page 2016
[4]

Brachmann, F

E. Brachmann, F. Michel, A. Krull, M. M. Yang, S. Gumhold, and C. Rother. Uncertainty-Driven 6D Pose Estimation of Objects and Scenes from a Single RGB Image. In Conference on Computer Vision and Pattern Recognition, 2016

work page 2016
[5]

M. Bui, S. Zakharov, S. Albarqouni, S. Ilic, and N. Navab. When Regression Meets Manifold Learning for Object Recognition and Pose Estimation. In International Confer- ence on Robotics and Automation, 2018

work page 2018
[6]

G. Cai, Y . Wang, M. Zhou, and L. He. Unsupervised Domain Adaptation with Adversarial Residual Transform Networks. In arXiv Preprint, 2018

work page 2018
[7]

Crivellaro, M

A. Crivellaro, M. Rad, Y . Verdie, K. M. Yi, P. Fua, and V . Lepetit. Robust 3D Object Tracking from Monocular Im- ages Using Stable Parts. IEEE Transactions on Pattern Anal- ysis and Machine Intelligence, 2018

work page 2018
[8]

T. Do, M. Cai, T. Pham, and I. Reid. Deep-6DPose: Recov- ering 6D Object Pose from a Single RGB Image. In arXiv Preprint, 2018

work page 2018

Show all 39 references

[9]

Drost, M

B. Drost, M. Ulrich, N. Navab, and S. Ilic. Model Globally, Match Locally: Efﬁcient and Robust 3D Object Recognition. In Conference on Computer Vision and Pattern Recognition, 2010

2010
[10]

Ganin, E

Y . Ganin, E. Ustinova, H. Ajakan, P. Germain, H. Larochelle, F. Laviolette, M. Marchand, and V . Lempitsky. Domain- Adversarial Training of Neural Networks. Journal of Ma- chine Learning Research, 2016

2016
[11]

I. J. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y . Bengio. Gen- erative Adversarial Nets. In Advances in Neural Information Processing Systems, 2014

2014
[12]

Gupta, J

S. Gupta, J. Hoffman, and J. Malik. Cross Modal Distillation for Supervision Transfer. In Conference on Computer Vision and Pattern Recognition, 2016

2016
[13]

Hartley and A

R. Hartley and A. Zisserman. Multiple View Geometry in Computer Vision. Cambridge University Press, 2000

2000
[14]

K. He, G. Gkioxari, P. Dollar, and R. Girshick. Mask R- CNN. In International Conference on Computer Vision , 2017

2017
[15]

Hinterstoisser, C

S. Hinterstoisser, C. Cagniart, S. Ilic, P. Sturm, N. Navab, P. Fua, and V . Lepetit. Gradient Response Maps for Real- Time Detection of Textureless Objects. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012

2012
[16]

Hinterstoisser, V

S. Hinterstoisser, V . Lepetit, P. Wohlhart, and K. Konolige. On Pre-Trained Image Features and Synthetic Images for Deep Learning. In European Conference on Computer Vi- sion Workshops, 2018

2018
[17]

Hodan, P

T. Hodan, P. Haluza, S. Obdrzalek, J. Matas, M. Lourakis, and X. Zabulis. T-LESS: An RGB-D Dataset for 6D Pose Estimation of Texture-less Objects. In IEEE Winter Confer- ence on Applications of Computer Vision, 2017. 6 Figure 6: Some qualitative results on Object #8 in Scene #...

2017
[18]

O. H. Jafari, S. K. Mustikovela, K. Pertsch, E. Brachmann, and C. Rother. IPose: Instance-Aware 6D Pose Estimation of Partly Occluded Objects. CoRR, abs/1712.01924, 2017

2017 arXiv
[19]

W. Kehl, F. Manhardt, F. Tombari, S. Ilic, and N. Navab. SSD-6D: Making RGB-Based 3D Detection and 6D Pose Estimation Great Again. In International Conference on Computer Vision, 2017

2017
[20]

Lee, H.-Y

H.-Y . Lee, H.-Y . Tseng, J.-B. Huang, M. Singh, and M.-H. Yang. Diverse Image-To-Image Translation via Disentan- gled Representations. In European Conference on Computer Vision, 2018

2018
[21]

W. Liu, D. Anguelov, D. Erhan, C. Szegedy, S. E. Reed, C.- Y . Fu, and A. C. Berg. SSD: Single Shot MultiBox Detector. In European Conference on Computer Vision, 2016

2016
[22]

M. Long, Y . Cao, J. Wang, and M. I. Jordan. Learning Trans- ferable Features with Deep Adaptation Networks. In Inter- national Conference on Machine Learning, 2015

2015
[23]

M ¨uller, F

F. M ¨uller, F. Bernard, O. Sotnychenko, D. Mehta, S. Sridhar, D. Casas, and C. Theobalt. Ganerated Hands for Real-Time 3D Hand Tracking from Monocular RGB. In Conference on Computer Vision and Pattern Recognition, 2018

2018
[24]

S. Peng, Y . Liu, Q. Huang, H. Bao, and X. Zhou. Pvnet: 7 Figure 9: Some qualitative results on Object #20 in Scene #13 of the T-LESS dataset. Figure 10: Some qualitative results on Object #20 in Scene #14 of the T-LESS dataset. Figure 11: Some qualitative results on Object #2...

2018 arXiv
[25]

Rad and V

M. Rad and V . Lepetit. BB8: A Scalable, Accurate, Robust to Partial Occlusion Method for Predicting the 3D Poses of Challenging Objects Without Using Depth. In International Conference on Computer Vision, 2017

2017
[26]

M. Rad, M. Oberweger, and V . Lepetit. Domain Transfer for 3D Pose Estimation from Color Images Without Manual An- notations. In Asian Conference on Computer Vision, 2018

2018
[27]

Redmon, S

J. Redmon, S. Divvala, R. Girshick, and A. Farhadi. You Only Look Once: Uniﬁed, Real-Time Object Detection. In Conference on Computer Vision and Pattern Recognition , 2016

2016
[28]

S. Ren, K. He, R. Girshick, and J. Sun. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. In Advances in Neural Information Processing Systems, 2015

2015
[29]

Rozantsev, M

A. Rozantsev, M. Salzmann, and P. Fua. Beyond Sharing Weights for Deep Domain Adaptation. In Conference on Computer Vision and Pattern Recognition, 2017

2017
[30]

Sundermeyer, Z

M. Sundermeyer, Z. Marton, M. Durner, M. Brucker, and R. Triebel. Implicit 3D Orientation Learning for 6D Object Detection from RGB Images. In European Conference on Computer Vision, 2018. 8

2018
[31]

Tekin, S

B. Tekin, S. N. Sinha, and P. Fua. Real-Time Seamless Single Shot 6D Object Pose Prediction. InConference on Computer Vision and Pattern Recognition, 2018

2018
[32]

Tobin, R

J. Tobin, R. Fong, A. Ray, J. Schneider, W. Zaremba, and P. Abbeel. Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World. In In- ternational Conference on Intelligent Robots and Systems , 2017

2017
[33]

Tzeng, J

E. Tzeng, J. Hoffman, T. Darrell, and K. Saenko. Simultane- ous Deep Transfer Across Domains and Tasks. In Interna- tional Conference on Computer Vision, 2015

2015
[34]

Wohlhart and V

P. Wohlhart and V . Lepetit. Learning Descriptors for Ob- ject Recognition and 3D Pose Estimation. In Conference on Computer Vision and Pattern Recognition, 2015

2015
[35]

Xiang, T

Y . Xiang, T. Schmidt, V . Narayanan, and D. Fox. PoseCNN: A Convolutional Neural Network for 6D Object Pose Esti- mation in Cluttered Scenes. Robotics: Science and Systems Conference, 2018

2018
[36]

Zakharov, W

S. Zakharov, W. Kehl, B. Planche, A. Hutter, and S. Ilic. 3D Object Instance Recognition and Pose Estimation Using Triplet Loss with Dynamic Margin. In International Confer- ence on Intelligent Robots and Systems, 2017

2017
[37]

Zakharov, B

S. Zakharov, B. Planche, Z. Wu, A. Hutter, H. Kosch, and S. Ilic. Keep It Unreal: Bridging the Realism Gap for 2.5D Recognition with Geometry Priors Only. In International Conference on 3D Vision, 2018

2018
[38]

Zakharov, I

S. Zakharov, I. Shugurov, and S. Ilic. DPOD: Dense 6D Pose Object Detector and Reﬁner. In International Conference on Computer Vision, 2019

2019
[39]

J.-Y . Zhu, T. Park, P. Isola, and A. A. Efros. Unpaired Image-To-Image Translation Using Cycle-Consistent Adver- sarial Networks. In International Conference on Computer Vision, 2017. 9

2017

Pith tools

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

[1] [1]

Balntas, A

V . Balntas, A. Doumanoglou, C. Sahin, J. Sock, R. Kousk- ouridas, and T.-K. Kim. Pose Guided RGBD Feature Learn- ing for 3D Object Pose Estimation. In International Confer- ence on Computer Vision, 2017

work page 2017

[2] [2]

Bousmalis, N

K. Bousmalis, N. Silberman, D. Dohan, D. Erhan, and D. Kr- ishnan. Unsupervised Pixel-Level Domain Adaptation with Generative Adversarial Networks. In Conference on Com- puter Vision and Pattern Recognition, 2017

work page 2017

[3] [3]

Bousmalis, G

K. Bousmalis, G. Trigeorgis, N. Silberman, D. Krishnan, and D. Erhan. Domain Separation Networks. In Advances in Neural Information Processing Systems , pages 343–351, 2016

work page 2016

[4] [4]

Brachmann, F

E. Brachmann, F. Michel, A. Krull, M. M. Yang, S. Gumhold, and C. Rother. Uncertainty-Driven 6D Pose Estimation of Objects and Scenes from a Single RGB Image. In Conference on Computer Vision and Pattern Recognition, 2016

work page 2016

[5] [5]

M. Bui, S. Zakharov, S. Albarqouni, S. Ilic, and N. Navab. When Regression Meets Manifold Learning for Object Recognition and Pose Estimation. In International Confer- ence on Robotics and Automation, 2018

work page 2018

[6] [6]

G. Cai, Y . Wang, M. Zhou, and L. He. Unsupervised Domain Adaptation with Adversarial Residual Transform Networks. In arXiv Preprint, 2018

work page 2018

[7] [7]

Crivellaro, M

A. Crivellaro, M. Rad, Y . Verdie, K. M. Yi, P. Fua, and V . Lepetit. Robust 3D Object Tracking from Monocular Im- ages Using Stable Parts. IEEE Transactions on Pattern Anal- ysis and Machine Intelligence, 2018

work page 2018

[8] [8]

T. Do, M. Cai, T. Pham, and I. Reid. Deep-6DPose: Recov- ering 6D Object Pose from a Single RGB Image. In arXiv Preprint, 2018

work page 2018