SEMPose: A Single End-to-end Network for Multi-object Pose Estimation

Dezong Zhao; Hao Wang; Shibei Xue; Xin Liu

REVIEW 4 major objections 7 minor 37 references

SEMPose: A Single End-to-end Network for Multi-object Pose Estimation

T0 review · 4 major / 7 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read SEMPose is a single end-to-end RGB-only network that estimates 6D poses for multiple objects at once, reporting top accuracy among compared single-model RGB methods on LM-O and YCB-V while running at 32 FPS with object-count-independent…

desk verdict Solid engineering paper with real benchmark gains, but the key occlusion-sampling trick rests on an unvalidated color assumption, and a few typos/omissions need fixing before I'd trust the details. read the letter →

arxiv 2411.14002 v1 pith:JY466DUM submitted 2024-11-21 cs.CV

Xin Liu , Hao Wang , Shibei Xue , Dezong Zhao This is my paper · ORCID

classification cs.CV

keywords 6Dposeestimationmulti-objectsceneend-to-endnetworkRGB-onlyocclusionhandlingvisiblepartsamplingfeaturepyramiddirectregression

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

SEMPose is a proposal for a single end-to-end convolutional network that estimates the full 6D pose of every object in an RGB image, without depth, 3D models, symmetry information, or ground-truth region proposals at inference. The paper reports that this one network, trained from RGB data only, reaches a mean ADD(-S) of 60.2% on the highly occluded LM-O dataset and an ADD-S AUC of 92.2% on the cluttered YCB-V dataset, both above the compared RGB-based single-model methods, and that its per-image inference time stays near 31 ms regardless of how many objects are present. A sympathetic reader would care because the result points toward a deployable real-time pose estimator for robotic grasping that needs only a camera, and because it shows that occlusion handling can be folded into training-sample selection rather than requiring a separate correspondence or refinement stage.

What carries the argument

The load-bearing mechanism is the combination of a Texture-Shape Guided Feature Pyramid Network (TS-FPN) and a visible-part positive sample selector. TS-FPN takes ResNet features and, at each level, concatenates upsampled low-frequency shape features with high-frequency texture features, applies channel-pooled attention to reweight both, and then rotates the feature maps to compute spatial attention along the height and width axes, producing five-scale fused maps that keep small and large objects learnable. The rotation head regresses a continuous 6D rotation representation and refines it with an initialization module and an iterative residual module that reuses early-layer features; the translation head separately regresses a projected center offset and a depth value, which are converted to a translation through the camera intrinsics. During training, positive samples are chosen by computing the minimum barrier distance in RGB color space between each feature-map cell and the ground-truth bounding-box boundary, under the assumption that boundary points lie outside the object, with cells whose normalized visibility probability exceeds 0.25 treated as positives.

What would settle it

Train or render an occlusion test set in which textureless objects are partially covered by occluders painted the same color as the object. If SEMPose's normalized visibility mask (Eq. 16) marks occluder pixels as visible and its mean ADD(-S) on LM-O falls to the level of the center-sampling ablation (row E0), then the visibility assumption carrying the occlusion gains is wrong.

Watch

Extended reading notes

Core claim

The paper's central claim is that jointly learned detection and pose regression can close the accuracy gap with indirect correspondence-plus-PnP methods if three design conditions hold: multi-scale features fuse texture and shape information, rotation and translation are regressed with separate iterative refinement heads rather than shared direct heads, and training positives are drawn from the visible parts of each object. In the reported experiments this combination gives SEMPose the best mean ADD(-S) among RGB-based single-model methods without refinement on LM-O (60.2%) and the best AUC of ADD-S on YCB-V (92.2%), and it reduces average translation error by 2.47 cm and average rotation error by 8.87 degrees relative to the GDR-Net baseline. The paper also claims the network sustains 32 FPS on a 640x480 image with runtime essentially flat as object count grows, because all objects are processed in the same forward pass.

Load-bearing premise

The visible-part sampling rests on assuming that ground-truth bounding-box boundary points always lie outside the target object and that the minimum color-barrier distance from any interior pixel to the boundary reliably separates visible object from occluder.

Editorial extensions

If this is right

A single shared model can replace per-object pose models, cutting deployment memory and training cost for scenes with many object categories.
A fixed 31 ms per-frame budget means pose estimates can be used in closed-loop robot control even as the number of objects in the scene grows.
End-to-end differentiability allows the pose head to be fine-tuned jointly with detection on new RGB-only data, without differentiable PnP or RANSAC.
The visible-part sampling rule converts occlusion handling into a segmentation-like training signal, so the method should degrade gracefully when the occluder and target have distinct colors.

Reading between the lines

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

A direct extension would test the color-based visibility assumption by adding synthetic occluders whose color closely matches the target; if ADD(-S) drops toward the center-sampling ablation, the gains are specific to color-separable occlusion rather than to visible-part sampling per se.
The constant-runtime claim would still require the detection head to keep finding all instances; an experimental stress test with more than ten heavily overlapping objects would show whether fixed FPS holds in extreme clutter.
The separation of rotation and translation losses may matter most for symmetric objects, where translation is well defined but rotation is not; an ablation restricted to symmetric YCB-V categories could reveal where the 2.47 cm and 8.87 degree gains come from.

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.

Desk Editor's Note

read the letter

You should know two things about SEMPose. First, it's a genuinely useful engineering result: a single RGB-only model that handles multiple objects in one forward pass, with runtime independent of object count and SOTA mean ADD(-S) on LM-O (60.2 vs GDR-Net's 56.1) and AUC of ADD-S 92.2 on YCB-V. Second, the main novelty that drives those gains—visible-part positive sampling—is both the biggest ablation driver (E0: 92.2 to 84.2) and the least rigorously validated component. The paper never quantitatively checks the visibility approximation against GT masks; Figure 5 shows one favorable example, and the per-object failures on Ape, Duck, and Cat are exactly the low-texture cases where that color-range barrier should fail. That doesn't kill the paper—the mean gains are still plausible—but it means the headline claim about occlusion handling is not as solid as the benchmark tables suggest.

What's actually new is the integration, not the pieces. TS-FPN, triplet attention, coordinate features, iterative refinement, and the sampling idea are all known and cited. The contribution is putting them together into one end-to-end network and showing it works; that's a legitimate contribution for an application-oriented venue. The ablation study is clean and the runtime analysis is honest. The paper also gives credit to the literature, including the authors' own earlier work, which is fine.

Soft spots, in order of severity. First, no code or weights and no error bars—single-run numbers on a few benchmarks are not enough to fully verify the claims. Second, the visibility sampling's assumptions are not validated; the paper should report IoU or accuracy against GT visibility masks on the training data, not just one qualitative image. Third, Eq. 4 defines e2 circularly (e2 = Phi(e1 × e2)), which is presumably a typo but needs a corrected re-derivation. Fourth, the paper says "for more parameter settings, please refer to [29]" without giving the threshold or alpha values in the text; those should be stated. Fifth, the relative weights of the training losses are omitted. None of these are fatal, but together they lower confidence.

This paper is for people building practical 6D pose estimators for robotics or studying end-to-end architectures, and for reviewers who care about benchmark progress with real-time constraints. It deserves a serious referee—the architecture and results are substantial enough to warrant the time—but the review should ask for code/weights, a corrected Eq. 4, and quantitative validation of the visibility approximation. I'd give it a conditional accept path rather than a desk reject.

Referee Report

4 major / 7 minor

Summary. The paper proposes SEMPose, a single end-to-end RGB-only network for multi-object 6D pose estimation. The method combines a texture-shape guided feature pyramid network (TS-FPN), decoupled iterative heads for rotation and translation, and a visibility-guided positive-sample selection scheme that uses a color-based minimum barrier distance to approximate visible object parts during training. The authors report state-of-the-art results among RGB-based single-model methods on LM-O (mean ADD(-S) 60.2% vs GDR-Net's 56.1%) and YCB-V (AUC of ADD-S 92.2% vs 89.1%), with an inference time of about 31 ms that is independent of the number of objects. The paper includes ablation studies on the feature fusion, pose regression strategy, head structure, and sampling strategy, plus a runtime analysis.

Significance. If the reported results hold, the paper offers a practical contribution: a real-time, single-model, RGB-only multi-object pose estimator that does not require depth, 3D models, or symmetry information at inference. The decoupled rotation/translation refinement and the visibility-guided sampling are plausible design ideas, and the benchmark comparisons use standard protocols and metrics. The main value is in the combination of these components into one end-to-end trainable system. However, the evidence is currently incomplete: the central visibility assumption is not quantitatively validated, the rotation construction in Eq. (4) contains a circular definition, and the training details are insufficiently specified for reproduction. The paper does not provide code, weights, or error bars, which matters because several headline margins over GDR-Net are modest and per-object results vary substantially.

major comments (4)

[Sec. 3.2, Eq. (4)] Equation (4) defines e2 = Φ(e1 × e2), which is self-referential: e2 appears on both sides, so the expression does not define a computable orthogonalization of the predicted vectors a1 and a2. As written, the rotation construction is undefined. This is a load-bearing step for the rotation head; please correct the formula (e.g., by Gram-Schmidt orthogonalization of a1 and a2) and re-derive the corresponding loss gradient.
[Sec. 3.4, Eqs. (14)–(16), Table 1, Table 4 row E0] The visibility-based positive sampling is the paper's main occlusion-handling contribution and drives a large ablation gain (row E0: AUC ADD-S drops from 92.2 to 84.2 when replaced by center sampling). Yet the core assumption—that boundary points of the GT bounding box lie outside the target object and that the minimum barrier distance in RGB color space separates visible pixels from occluders—is not validated quantitatively. Figure 5 shows only a single favorable example, and no IoU or accuracy statistics against ground-truth visibility masks are reported. The failure mode is visible in Table 1: on LM-O, SEMPose underperforms GDR-Net on Ape (33.2 vs 44.9), Duck (29.0 vs 40.0), Cat (30.1 vs 30.6), and Glue (67.3 vs 73.7), which are low-texture objects where color-range barriers should be least reliable. Please report per-object visibility-mask agreement statistics and analyze how sampling quality correlates with per-object ADD(-S) results.
[Sec. 4.1 and Sec. 3.4] The training objective is not fully specified. The paper defines L_rot_6d, L_tran1, L_tran2, plus classification and bounding-box losses, but does not state how these losses are weighted in the total loss or how the four heads' losses are combined. In addition, the positive-sample parameters (e.g., number of boundary points m, cell size, and any additional thresholds) are deferred to [29] with 'for more parameter settings, please refer to [29]'. This makes the reported ablations (B0, C0–C2, D0, E0) irreproducible from the manuscript alone. Please report the complete loss function and all sampling hyperparameters.
[Tables 1–4] All comparisons and ablations are based on a single training run, with no standard deviations or multiple seeds. The headline margins over GDR-Net are modest (60.2 vs 56.1 on LM-O; 92.2 vs 89.1 AUC ADD-S on YCB-V), and per-object results are highly variable. The claim that SEMPose 'outperforms other RGB-based single-model methods' would be strengthened by reporting variance across at least a few seeds, or a clear statement of the run-to-run stability. Please also clarify which numbers in Table 3 are recomputed by the authors and which are taken from the original publications.

minor comments (7)

[Sec. 3.4, Eq. (14)] The notation is inconsistent: the left-hand side uses V(p,b), but the right-hand side minimizes over b∈B, and the Euclidean distance term is written as d(p,d) instead of d(p,b). Please make the definition of V(p) (or V(p,b)) explicit and correct the distance notation.
[Sec. 3.4, Eq. (15)] The '3 max i=1' notation is unclear; it should be written as max over i = 1,2,3 to indicate the maximum over the three RGB channels.
[Sec. 4.1, Eq. (18)] The ADD-S formula uses the same model point x on both the ground-truth and predicted sides, with the minimum taken over x2 but not used in the expression. The standard symmetric metric should be avg_{x1} min_{x2} ||(R_gt x1 + t_gt) - (R_pred x2 + t_pred)||. Please correct the equation to match the metric used in the tables.
[Sec. 4.3] The phrase 'In Table. In row B0' should read 'In Table 4, row B0'.
[Table 3 caption] The citation for SilhoNet appears as 'SILHONET[]' with an empty bracket; please fill in the reference.
[Sec. 4.4] Figure 7 shows that runtime is independent of object count, but no numerical runtime table for competing methods under the same hardware is provided. Please include the numbers behind Figure 7 or a table with inference times.
[References] References [36] and [37] are generic surveys and are not needed for the specific claims; consider replacing them with more targeted citations on object detection or pose estimation.

Circularity Check

1 steps flagged · score 2.0 of 10

No load-bearing circularity: benchmark results are externally validated and self-citations [23,24] are related-work only; the sole self-referential step is Eq. (4), where e2 is defined as Phi(e1 x e2) in terms of itself.

self definitional [Sec. 3.2, Eq. (4), rotation regression strategy]
"Next, we construct three standard orthogonal bases 𝒆1,𝒆2,and𝒆3 from these two 3D vectors. ... This process can be represented as: 𝒆1= 𝛷(𝒂1), 𝒆2=𝛷(𝒆1× 𝒆2), 𝒆3= 𝒆1× 𝒆2, (4) where 𝛷(⋅) denotes vector normalization, while the symbol × denotes the vector cross product."

As printed, the second basis vector is defined through itself: 𝒆2 = Φ(𝒆1 × 𝒆2) has 𝒆2 on both sides, and 𝒆1 × 𝒆2 is always orthogonal to 𝒆2, so no nonzero 𝒆2 satisfies the equation. The derivation step from the regressed 6-D vector to the rotation basis is therefore self-referential and ill-defined as written; the standard Gram-Schmidt 6D construction (cited [22]) instead orthogonalizes 𝒂2 against 𝒆1 and then sets 𝒆3 = 𝒆1 × 𝒆2. The flaw is internal to the rotation-head construction and does not reduce the externally measured ADD(-S) benchmark figures to their inputs, so it carries limited weight in the score.

full rationale

The central claims—ADD(-S) 60.2% on LM-O and AUC of ADD-S 92.2% on YCB-V, plus runtime near 31 ms—are measured against external BOP benchmarks under standard metrics; no quantity is fitted to the test targets, and no ablation predicts a value that entered the training process. The visible-part positive sampling (Sec. 3.4, Eqs. 14–16) is a fixed training-time heuristic using GT bounding boxes and the external minimum-barrier distance [31], adapted from Hai et al. [29]; its ablation row E0 is an internal comparison, not a renamed fit. The self-citations [23, 24] (Guan et al., sharing the corresponding author) appear only in the related-work survey of indirect correspondence methods and are not load-bearing for any SEMPose design decision or accuracy claim, so per the review rules they do not raise the score. The sampling-parameter citation [29] is external. The one genuine self-referential step is Eq. (4), where e2 is defined via e1 × e2, i.e., through itself; as printed that step is ill-defined, but it is an internal Gram-Schmidt-style step whose standard form is well known and it does not affect the externally validated comparisons. The visibility-model assumptions (boundary points outside the object; color barrier separates occluders) are a robustness and correctness risk, reflected in per-object LM-O drops on low-texture objects; this is not a circularity. Score 2 reflects one minor self-referential derivation step plus non-load-bearing self-citations, with the core results self-contained against external benchmarks.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The central performance claim rests on standard deep learning assumptions plus a few hand-set constants. The visibility model and loss weights are the least constrained pieces; the paper does not state the total loss composition or report sensitivity to alpha and the positive-sample threshold.

free parameters (3)

alpha (color-distance balance in visibility scoring) = 0.1
Hand-set in Eq. 14; no sensitivity analysis. It controls how much RGB brightness differences weigh against pixel distance in the positive-sample scoring.
positive-sample visibility threshold = 0.25
Cells with P(c) > 0.25 are positive samples (Sec. 3.4); hand-set and the paper refers to [29] for other parameter settings.
relative weights of the training losses = not reported
The final loss combines rotation, translation, classification, and box losses, but the mixing coefficients are not stated. Reproducing training requires them.

assumptions (5)

standard math The 6D rotation representation is continuous enough for direct regression and can be orthogonalized into a rotation matrix.
Used in Sec. 3.2, following Zhou et al. [22].
domain assumption Minimum barrier distance computed on RGB brightness approximates the visible part of an object.
Used in Eqs. 14-16 and Sec. 3.4, inherited from [29, 31].
domain assumption The camera intrinsic matrix K is known and fixed.
Eq. 7 converts predicted offsets and depth into translation; the datasets provide intrinsics.
domain assumption Appending coordinate features makes translation regression easier because convolutions are translation-variant only through padding.
Invoked in Sec. 3.3, citing CoordConv [17].
domain assumption Weight sharing across feature pyramid scales is beneficial when positive samples are sparse.
Stated in Sec. 3.3, following FCOS [15].

how reviews work

0 comments

Cite this review

Pith. "Pith review of SEMPose: A Single End-to-end Network for Multi-object Pose Estimation." pith.science (2026). https://pith.science/paper/JY466DUM

@misc{pith2026241114002,
  author       = {Pith},
  title        = {Pith review of: SEMPose: A Single End-to-end Network for Multi-object Pose Estimation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JY466DUM}},
  note         = {Machine review of arXiv:2411.14002}
}

read the original abstract

In computer vision, estimating the six-degree-of-freedom pose from an RGB image is a fundamental task. However, this task becomes highly challenging in multi-object scenes. Currently, the best methods typically employ an indirect strategy, which identifies 2D and 3D correspondences, and then solves with the Perspective-n-Points method. Yet, this approach cannot be trained end-to-end. Direct methods, on the other hand, suffer from lower accuracy due to challenges such as varying object sizes and occlusions. To address these issues, we propose SEMPose, an end-to-end multi-object pose estimation network. SEMPose utilizes a well-designed texture-shape guided feature pyramid network, effectively tackling the challenge of object size variations. Additionally, it employs an iterative refinement head structure, progressively regressing rotation and translation separately to enhance estimation accuracy. During training, we alleviate the impact of occlusion by selecting positive samples from visible parts. Experimental results demonstrate that SEMPose can perform inference at 32 FPS without requiring inputs other than the RGB image. It can accurately estimate the poses of multiple objects in real time, with inference time unaffected by the number of target objects. On the LM-O and YCB-V datasets, our method outperforms other RGB-based single-model methods, achieving higher accuracy. Even when compared with multi-model methods and approaches that use additional refinement, our results remain competitive.

Figures

Figures reproduced from arXiv: 2411.14002 by the authors.

**Figure 1.** Framework of SEMPose. Given an input RGB image I, SEMPose first uses Resnet50 to extract basic features. Then, SEMPose employs a Texture-Shape Guided Feature Pyramid Network (TS-FPN) to fuse these features and produce feature maps at five different scales. Finally, SEMPose uses four heads to predict the categories, bounding boxes, rotations, and translations of the objects. After post-processing, the poses are obtai… view at source ↗

**Figure 2.** The structure diagram of texture shape feature fusion module. generate attention weights. These attention weights are applied to  𝑖 𝑙 and  𝑖 ℎ separately, and then the weighted highfrequency/low-frequency features are added back to  𝑖 𝑙 / 𝑖 ℎ to obtain enhanced features ̂ 𝑖 ℎ /̂ 𝑖 𝑙 . This enhancement process can be described as: ̂ 𝑖 ℎ =  𝑖 ℎ ⊕ ( 𝑖 𝑙 ⊗ 𝜎(𝐶𝑜𝑛𝑣(𝐶-𝑃 𝑜𝑜𝑙(𝐶( 𝑖 ℎ ,  𝑖 𝑙 ))))), ̂ 𝑖 𝑙 =  𝑖 𝑙 … view at source ↗

**Figure 3.** The structure of the rotation head. Therefore, we did not directly regress the translation vector 𝐭 = ( 𝑡𝑥 , 𝑡𝑦 , 𝑡𝑧 )𝑇 like in PoET. First, the transformation between the camera coordinate system and the pixel coordinate system can be expressed as 𝑡𝑧 ( 𝑐𝑥 , 𝑐𝑦 , 1 )𝑇 = 𝐾 ( 𝑡𝑥 , 𝑡𝑦 , 𝑡𝑧 )𝑇 , where 𝑐𝑥 and 𝑐𝑦 represent the coordinates of the projected 3D point on the image, and K represents the camera intrinsic param… view at source ↗

Figures from the paper (4 more)

**Figure 4.** Figure 4: Schematic diagrams of the three different sampling strategies [15, 16, 29]. A darker green color indicates a higher probability of sampling [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]

**Figure 5.** Figure 5: The schematic diagram of the sampling process from visible parts. 𝒓6𝑑 = 𝒓𝑖𝑛𝑖𝑡 + Δ𝒓. (13) Our translation head and rotation head are essentially the same. The only difference is that we use two branches in the translation head, as shown in [PITH_FULL_IMAGE:figures/full…

**Figure 6.** Figure 6: Qualitative comparison results on YCB-V [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]

**Figure 7.** Figure 7: Runtime comparison under the same conditions. The running time of existing methods becomes longer as the number of objects to be estimated increases[2, 4], whereas our method is unaffected. methods using only RGB images, SEMPose achieves stateof-the-art performance on…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

37 extracted references · 36 canonical work pages

[29]

Rigidity-awaredetec- tion for 6d object pose estimation,

Y.Hai,R.Song,J.Li,M.Salzmann,andY.Hu,“Rigidity-awaredetec- tion for 6d object pose estimation,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 8927–8936

work page 2023
[1]

Deep learning-based pose prediction for visual servoing of robotic manipulators using image similarity,

Y. He, J. Gao, and Y. Chen, “Deep learning-based pose prediction for visual servoing of robotic manipulators using image similarity,” inNeurocomputing, vol. 491. Elsevier, 2022, pp. 343–352

work page 2022
[2]

Cdpn: Coordinates-based disentangled posenetworkforreal-timergb-based6-dofobjectposeestimation,

Z. Li, G. Wang, and X. Ji, “Cdpn: Coordinates-based disentangled posenetworkforreal-timergb-based6-dofobjectposeestimation,”in ProceedingsoftheIEEE/CVFInternationalConferenceonComputer Vision, 2019, pp. 7678–7687

work page 2019
[3]

Deepresiduallearningforimage recognition,

K.He,X.Zhang,S.Ren,andJ.Sun,“Deepresiduallearningforimage recognition,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 770–778

work page 2016
[4]

Cosypose:Consistent multi-view multi-object 6d pose estimation,

Y.Labbé,J.Carpentier,M.Aubry,andJ.Sivic,“Cosypose:Consistent multi-view multi-object 6d pose estimation,” inComputer Vision– ECCV 2020: 16th European Conference, Glasgow, UK, August 23– 28, 2020, Proceedings, Part XVII 16. Springer, 2020, pp. 574–591

work page 2020
[5]

Gdr-net: Geometry- guided direct regression network for monocular 6d object pose esti- mation,

G. Wang, F. Manhardt, F. Tombari, and X. Ji, “Gdr-net: Geometry- guided direct regression network for monocular 6d object pose esti- mation,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 16611–16621

work page 2021
[6]

Feature pyramid networks for object detection,

T.-Y.Lin,P.Dollár,R.Girshick,K.He,B.Hariharan,andS.Belongie, “Feature pyramid networks for object detection,” inProceedings of the IEEE conference on computer vision and pattern recognition, 2017, pp. 2117–2125

work page 2017
[7]

Posecnn: a con- volutional neural network for 6d object pose estimation in cluttered scenes,

Y. Xiang, T. Schmidt, V. Narayanan, and D. Fox, “Posecnn: a con- volutional neural network for 6d object pose estimation in cluttered scenes,”Robotics: Science and Systems (RSS),, 2018

work page 2018

Show all 37 references

[8]

Pvnet: Pixel- wise voting network for 6dof pose estimation,

S. Peng, Y. Liu, Q. Huang, X. Zhou, and H. Bao, “Pvnet: Pixel- wise voting network for 6dof pose estimation,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 4561–4570

2019
[9]

Poet: pose estimation transformer for single-view, multi-object 6d pose estimation,

T.G.Jantos,M.A.Hamdad,W.Granig,S.Weiss,andJ.Steinbrener, “Poet: pose estimation transformer for single-view, multi-object 6d pose estimation,” inConference on Robot Learning. PMLR, 2023, pp. 1060–1070

2023
[10]

Silhonet: An rgb method for 6d object pose estimation,

G. Billings and M. Johnson-Roberson, “Silhonet: An rgb method for 6d object pose estimation,”IEEE Robotics and Automation Letters, vol. 4, no. 4, pp. 3727–3734, 2019

2019
[11]

Segmentation-driven6d object pose estimation,

Y.Hu,J.Hugonot,P.Fua,andM.Salzmann,“Segmentation-driven6d object pose estimation,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 3385–3394

2019
[12]

Dpod: 6d pose object detector and refiner,

S. Zakharov, I. Shugurov, and S. Ilic, “Dpod: 6d pose object detector and refiner,” inProceedings of the IEEE/CVF international confer- ence on computer vision, 2019, pp. 1941–1950

2019
[13]

Deepim:Deepiterative matching for 6d pose estimation,

Y.Li,G.Wang,X.Ji,Y.Xiang,andD.Fox,“Deepim:Deepiterative matching for 6d pose estimation,” inProceedings of the European Conference on Computer Vision (ECCV), 2018, pp. 683–698

2018
[14]

Rnnpose: 6-dof object pose estimation via recurrent correspondence field estimation and pose optimization,

Y. Xu, K.-Y. Lin, G. Zhang, X. Wang, and H. Li, “Rnnpose: 6-dof object pose estimation via recurrent correspondence field estimation and pose optimization,”IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024

2024
[15]

Fcos: a simple and strong anchor-free object detector,

Z. Tian, C. Shen, H. Chen, and T. He, “Fcos: a simple and strong anchor-free object detector,”IEEE Transactions on Pattern Analysis & Machine Intelligence, vol. 44, no. 04, pp. 1922–1933, 2022

1922
[16]

Bridging the gap betweenanchor-basedandanchor-freedetectionviaadaptivetraining sample selection,

S. Zhang, C. Chi, Y. Yao, Z. Lei, and S. Z. Li, “Bridging the gap betweenanchor-basedandanchor-freedetectionviaadaptivetraining sample selection,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 9759–9768

2020
[17]

An intriguing failing of convolutional neural net- works and the coordconv solution,

R. Liu, J. Lehman, P. Molino, F. Petroski Such, E. Frank, A. Sergeev, and J. Yosinski, “An intriguing failing of convolutional neural net- works and the coordconv solution,”Advances in neural information processing systems, vol. 31, 2018

2018
[18]

Super-convergence: very fast training of neural networks using large learning rates,

L. N. Smith and N. Topin, “Super-convergence: very fast training of neural networks using large learning rates,” inArtificial intelligence and machine learning for multi-domain operations applications, vol. 11006. SPIE, 2019, pp. 369–386

2019
[19]

Learning 6d object pose estimation using 3d object coordinates,

E. Brachmann, A. Krull, F. Michel, S. Gumhold, J. Shotton, and C. Rother, “Learning 6d object pose estimation using 3d object coordinates,” inComputer Vision–ECCV 2014: 13th European Con- ference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part II 13. Springer, 20...

2014
[20]

3dposeregressionusingconvo- lutional neural networks,

S.Mahendran,H.Ali,andR.Vidal,“3dposeregressionusingconvo- lutional neural networks,” inProceedings of the IEEE International Conference on Computer Vision Workshops, 2017, pp. 2174–2182

2017
[21]

Efficientpose: An efficient, accurate and scalableend-to-end6dmultiobjectposeestimationapproach,

Y. Bukschat and M. Vetter, “Efficientpose: An efficient, accurate and scalableend-to-end6dmultiobjectposeestimationapproach,” arXiv preprint arXiv:2011.04307, 2020. Xin Liu et al.:Preprint submitted to Elsevier Page 10 of 11 SEMPose: A Single End-to-end Network for Multi-objec...

2011 arXiv
[22]

On the continuity of rotation representations in neural networks,

Y. Zhou, C. Barnes, J. Lu, J. Yang, and H. Li, “On the continuity of rotation representations in neural networks,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 5745–5753

2019
[23]

Hrpose: Real-time high-resolution 6d pose estimation network using knowledge distillation,

Q. Guan, Z. Sheng, and S. Xue, “Hrpose: Real-time high-resolution 6d pose estimation network using knowledge distillation,”Chinese Journal of Electronics, vol. 32, no. 1, pp. 189–198, 2023

2023
[24]

High-resolution representation object pose estimation from monocular images,

Q. Guan, W. Li, S. Xue, and D. Li, “High-resolution representation object pose estimation from monocular images,” in2021 China Au- tomation Congress (CAC). IEEE, 2021, pp. 980–984

2021
[25]

Rotate to attend: Convolutional triplet attention module,

D. Misra, T. Nalamada, A. U. Arasanipalai, and Q. Hou, “Rotate to attend: Convolutional triplet attention module,” inProceedings of the IEEE/CVF winter conference on applications of computer vision, 2021, pp. 3139–3148

2021
[26]

Rethinking the necessity of image fusion in high-level vision tasks: A practical infrared and visibleimagefusionnetworkbasedonprogressivesemanticinjection and scene fidelity,

L. Tang, H. Zhang, H. Xu, and J. Ma, “Rethinking the necessity of image fusion in high-level vision tasks: A practical infrared and visibleimagefusionnetworkbasedonprogressivesemanticinjection and scene fidelity,”Information Fusion, vol. 99, p. 101870, 2023

2023
[27]

Bop challenge 2022 on detec- tion, segmentation and pose estimation of specific rigid objects,

M. Sundermeyer, T. Hodaň, Y. Labbe, G. Wang, E. Brachmann, B. Drost, C. Rother, and J. Matas, “Bop challenge 2022 on detec- tion, segmentation and pose estimation of specific rigid objects,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, ...

2022
[28]

Blenderproc: Reducing the reality gap with photorealistic render- ing,

M.Denninger,M.Sundermeyer,D.Winkelbauer,D.Olefir,T.Hodan, Y. Zidan, M. Elbadrawy, M. Knauer, H. Katam, and A. Lodhi, “Blenderproc: Reducing the reality gap with photorealistic render- ing,” inInternational Conference on Robotics: Sciene and Systems, RSS 2020, 2020

2020
[30]

Open chal- lenges for monocular single-shot 6d object pose estimation,

S. Thalhammer, P. Hönig, J.-B. Weibel, and M. Vincze, “Open chal- lenges for monocular single-shot 6d object pose estimation,”arXiv preprint arXiv:2302.11827, 2023

2023 arXiv
[31]

Minimum barrier superpixel segmentation,

Y. Hu, Y. Li, R. Song, P. Rao, and Y. Wang, “Minimum barrier superpixel segmentation,”Image and Vision Computing, vol. 70, pp. 1–10, 2018

2018
[32]

Epos: Estimating 6d pose of ob- jects with symmetries,

T. Hodan, D. Barath, and J. Matas, “Epos: Estimating 6d pose of ob- jects with symmetries,” inProceedings of the IEEE/CVF conference oncomputervisionandpatternrecognition ,2020,pp.11703–11712

2020
[33]

Sam-6d: Segment anything model meets zero-shot 6d object pose estimation,

J. Lin, L. Liu, D. Lu, and K. Jia, “Sam-6d: Segment anything model meets zero-shot 6d object pose estimation,” inProceedings of the IEEE/CVFConferenceonComputerVisionandPatternRecognition , 2024, pp. 27906–27916

2024
[34]

Hipose: Hierarchical binary surface encoding and correspondence pruning for rgb-d 6dof objectposeestimation,

Y.Lin,Y.Su,P.Nathan,S.Inuganti,Y.Di,M.Sundermeyer,F.Man- hardt, D. Stricker, J. Rambach, and Y. Zhang, “Hipose: Hierarchical binary surface encoding and correspondence pruning for rgb-d 6dof objectposeestimation,”in ProceedingsoftheIEEE/CVFConference on Computer Vision and Pat...

2024
[35]

Megapose: 6d poseestimationofnovelobjectsviarender&compare,

Y.Labbé,L.Manuelli,A.Mousavian,S.Tyree,S.Birchfield,J.Trem- blay, J. Carpentier, M. Aubry, D. Fox, and J. Sivic, “Megapose: 6d poseestimationofnovelobjectsviarender&compare,”in Conference on Robot Learning. PMLR, 2023, pp. 715–725

2023
[36]

A gentle introduction to object recognition with deep learning,

J. Brownlee, “A gentle introduction to object recognition with deep learning,”Machine Learning Mastery, vol. 5, p. 10, 2019

2019
[37]

Object detection using deep learn- ing, cnns and vision transformers: a review,

A. B. Amjoud and M. Amrouch, “Object detection using deep learn- ing, cnns and vision transformers: a review,”IEEE Access, 2023. Xin Liu et al.:Preprint submitted to Elsevier Page 11 of 11

2023

Pith tools

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

[1] [29]

Rigidity-awaredetec- tion for 6d object pose estimation,

Y.Hai,R.Song,J.Li,M.Salzmann,andY.Hu,“Rigidity-awaredetec- tion for 6d object pose estimation,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 8927–8936

work page 2023

[2] [1]

Deep learning-based pose prediction for visual servoing of robotic manipulators using image similarity,

Y. He, J. Gao, and Y. Chen, “Deep learning-based pose prediction for visual servoing of robotic manipulators using image similarity,” inNeurocomputing, vol. 491. Elsevier, 2022, pp. 343–352

work page 2022

[3] [2]

Cdpn: Coordinates-based disentangled posenetworkforreal-timergb-based6-dofobjectposeestimation,

Z. Li, G. Wang, and X. Ji, “Cdpn: Coordinates-based disentangled posenetworkforreal-timergb-based6-dofobjectposeestimation,”in ProceedingsoftheIEEE/CVFInternationalConferenceonComputer Vision, 2019, pp. 7678–7687

work page 2019

[4] [3]

Deepresiduallearningforimage recognition,

K.He,X.Zhang,S.Ren,andJ.Sun,“Deepresiduallearningforimage recognition,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 770–778

work page 2016

[5] [4]

Cosypose:Consistent multi-view multi-object 6d pose estimation,

Y.Labbé,J.Carpentier,M.Aubry,andJ.Sivic,“Cosypose:Consistent multi-view multi-object 6d pose estimation,” inComputer Vision– ECCV 2020: 16th European Conference, Glasgow, UK, August 23– 28, 2020, Proceedings, Part XVII 16. Springer, 2020, pp. 574–591

work page 2020

[6] [5]

Gdr-net: Geometry- guided direct regression network for monocular 6d object pose esti- mation,

G. Wang, F. Manhardt, F. Tombari, and X. Ji, “Gdr-net: Geometry- guided direct regression network for monocular 6d object pose esti- mation,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 16611–16621

work page 2021

[7] [6]

Feature pyramid networks for object detection,

T.-Y.Lin,P.Dollár,R.Girshick,K.He,B.Hariharan,andS.Belongie, “Feature pyramid networks for object detection,” inProceedings of the IEEE conference on computer vision and pattern recognition, 2017, pp. 2117–2125

work page 2017

[8] [7]

Posecnn: a con- volutional neural network for 6d object pose estimation in cluttered scenes,

Y. Xiang, T. Schmidt, V. Narayanan, and D. Fox, “Posecnn: a con- volutional neural network for 6d object pose estimation in cluttered scenes,”Robotics: Science and Systems (RSS),, 2018

work page 2018