Pith. sign in

REVIEW 5 major objections 5 minor 64 references

3DGeoDet: General-purpose Geometry-aware Image-based 3D Object Detection

T0 review · 5 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read 3DGeoDet claims that two depth-derived cues — voxel occupancy scores and a TSDF volume — embedded in a voxel feature volume lift RGB-only 3D detection by up to 9.3 mAP@0.5 on indoor and outdoor benchmarks.

desk verdict A solid engineering combination of depth-derived occupancy and TSDF cues that beats prior work by a wide margin, but the 'no 3D supervision' framing is overstated and the depth head itself is never isolated. read the letter →

arxiv 2506.09541 v1 pith:MGGXZO3X submitted 2025-06-11 cs.CV

Yi Zhang , Yi Wang , Yawen Cui , Lap-Pui Chau This is my paper · ORCID
classification cs.CV
keywords 3Dobjectdetectionmonocularmulti-viewvoxeloccupancyattentionTSDFshapinggeometry-awarepredicteddepthimage-based
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

Image-based 3D object detection is hampered by the absence of geometric cues: when 2D features are back-projected into voxels, the network cannot tell empty space from occupied surfaces. 3DGeoDet attacks this by predicting a depth map for each RGB image and converting that depth into two cheap 3D signals — per-voxel occupancy scores and a truncated signed distance function (TSDF) volume — which are multiplied into and concatenated with the image-derived 3D feature volume before detection. The paper claims this geometry-aware volume yields state-of-the-art results from RGB alone: 9.3 mAP@0.5 over prior image-based methods on SUN RGB-D, 3.3 on ScanNetV2, and 0.19 AP3D@0.7 on KITTI. It also claims strong data efficiency, matching a 70-view method with 20 views, and end-to-end training without ground-truth point clouds or TSDF supervision, with the lightweight depth head adding an L1 depth loss. If true, the approach makes monocular and few-view 3D detection substantially more practical for indoor and outdoor settings.

What carries the argument

The load-bearing mechanisms are two modules that consume the same predicted depth maps. Voxel Occupancy Attention (the explicit cue) projects each predicted depth map into a sparse point cloud, computes each voxel's occupancy score as the fraction of points inside it, aggregates scores over views, and element-wise multiplies them into the 3D feature volume, making occupied regions stand out. TSDF Shaping (the implicit cue) fuses the depth maps into a truncated signed distance function volume using the standard weighted TSDF fusion rule, then concatenates this volume with the occupancy-weighted feature volume after a 3D convolution, giving each voxel a signed distance-to-nearest-surface signal. Together they rewrite the voxelized feature volume so the detection head receives both "is this region occupied?" and "how far is the nearest surface?" information; end-to-end training with bounding-box supervision plus an L1 depth loss ties the depth head to the detection task.

What would settle it

A decisive check would be to hold the trained detector fixed and swap the predicted depth maps at inference for ground-truth depth on one benchmark and for randomized depth on another: if the gains over the no-geometry baseline do not grow with accurate depth or survive corrupted depth, the occupancy and TSDF cues are not carrying the reported improvement.

Watch

Extended reading notes

Core claim

On the paper's terms, 3DGeoDet establishes that predicted depth is enough of a geometric bridge to make an image-to-voxel detector see 3D structure. The detector starts from a voxelized feature volume built by back-projecting 2D features from one or more posed images. A lightweight depth head produces per-image depth maps; Voxel Occupancy Attention counts how many back-projected depth points fall inside each voxel and multiplies the volume by these occupancy scores, while TSDF Shaping runs standard TSDF fusion over the same depth maps and concatenates the resulting distance-to-surface volume with the feature volume. The refined volume feeds a 3D detection head. The paper reports consistent gains over state-of-the-art image-based methods across all three benchmarks — larger on indoor single-view (SUN RGB-D) and multi-view (ScanNetV2), smaller but present outdoors (KITTI) — and interprets this as evidence that explicit and implicit geometric cues jointly disambiguate correspondences between image features and 3D space.

Load-bearing premise

The load-bearing premise is that the predicted depth maps are reliable enough that both geometric modules — the occupied-voxel scoring and the distance-to-surface volume — point at the right places; if depth estimates are poor, those cues mislead the detector, and the depth head itself needs ground-truth depth labels during training.

Editorial extensions

If this is right

  • On ScanNetV2, 20 input views match or beat the prior method's 70-view performance, so the geometric cues reduce the number of views needed for reliable indoor detection.
  • On SUN RGB-D, the reported 9.3 mAP@0.5 gain over the single-view baseline narrows the gap between RGB-only and RGB-D or point-cloud indoor detection.
  • On KITTI, the 0.19 AP3D@0.7 improvement at the moderate level shows the same modules transfer to monocular outdoor car detection.
  • Because the method trains end-to-end with bounding-box supervision rather than ground-truth TSDF or LiDAR data, applying it to a new scene type requires only RGB images plus box labels and depth supervision for the depth head.

Reading between the lines

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

  • An untested extension is to freeze a pretrained depth network and retrain only the geometry modules; if the reported gains persist, the co-trained depth head is not the source of the improvement.
  • The same occupancy-and-TSDF shaping recipe could likely attach to other voxelized or bird's-eye-view detectors, since it only rewrites a feature volume.
  • A natural next step the authors mention is replacing the TSDF volume with richer implicit representations such as 3D Gaussian splatting, which would likely tighten the geometric cue further.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 5 minor

Summary. 3DGeoDet proposes to improve image-based 3D object detection by injecting geometry cues derived from a predicted depth map into a voxel feature volume. A depth head predicts depth; Algorithm 1 computes per-voxel occupancy scores by back-projecting depth into sparse point clouds, and Algorithm 2 fuses the predicted depth into a TSDF volume. The voxel feature volume is multiplied by occupancy scores and concatenated with the TSDF volume, then processed by ImVoxelNet-style indoor/outdoor detection heads. The model is trained end-to-end with a detection loss plus an L1 depth loss (Eq. 8). Experiments on ScanNetV2, SUN RGB-D, and KITTI report state-of-the-art results over image-based methods, including a 9.3 mAP@0.5 gain on SUN RGB-D, a 3.3 mAP@0.5 gain on ScanNetV2, and a 0.19 AP3D@0.7 gain on KITTI moderate. The paper emphasizes generality across single/multi-view and indoor/outdoor settings and stresses that the method does not require supervision from 3D signals.

Significance. If the results hold, the paper contributes a simple, effective way to make image-to-voxel detectors geometry-aware without requiring LiDAR point clouds or TSDF ground truth, and it demonstrates generality from multi-view indoor detection to monocular outdoor detection. The controlled comparison in Table II (20-view training, varying inference views) shows large and consistent gains over CN-RMA, which is a credible positive result. The paper also provides explicit algorithms and clear architectural details. However, the headline claim that the method works 'without requiring supervision from 3D signals' is not supported as written, because the depth head is supervised by ground-truth depth maps and the ablation baseline is also trained with ground-truth depth supervision. The magnitude of the reported gains may therefore depend on a form of 3D supervision that is never isolated or ablated. The significance is conditional on resolving this supervision issue and adding a sensitivity analysis of the depth signal.

major comments (5)
  1. [Abstract; §III-E Eq. (8); §IV-C.1] The abstract states that 3DGeoDet works 'without requiring supervision from 3D signals,' but the training loss in Eq. (8) explicitly includes λ L_depth, an L1 depth loss, and Section IV-C.1 states that even the ablation baseline is 'also trained using ground truth depth maps as supervision.' Ground-truth depth maps are 3D geometric supervision, and the occupancy scores (Algorithm 1) and TSDF values (Algorithm 2) are both computed from the depth head's output. This contradiction is load-bearing for the central claim. Please either retrain and evaluate the model with λ=0 (removing the depth loss entirely) and show that the reported gains persist, or revise the abstract and framing to acknowledge that depth supervision is required.
  2. [§III-C, §III-D, §III-E; Tables V–VII] No experiment measures the quality of the predicted depth or the sensitivity of the proposed modules to depth errors. The gain of the occupancy and TSDF modules is shown only when the depth head is trained with ground-truth depth. To support the claim that the pipeline is robust and that the improvements come from geometry cues rather than from depth supervision itself, report depth accuracy (e.g., Abs Rel, Sq Rel, δ1) on the three benchmarks and run an oracle experiment at inference with ground-truth depth maps replacing predicted depth. If the oracle gain is much larger than the predicted-depth gain, depth quality is a bottleneck that should be discussed; if the gain is similar, the depth head is sufficient.
  3. [Algorithm 2, line 5] In Algorithm 2, line 5, the signed distance is written as sd f_j^i = ||C_j - R_i|| - D_i(I_ij), where R_i was defined as the 4x4 camera extrinsic matrix. The norm of a matrix is not defined in this context, so the formula cannot be implemented as written. The standard TSDF fusion uses the Euclidean distance from the voxel center to the camera center; please replace R_i by the camera center t_i extracted from the extrinsic matrix and likewise correct the weight w_j^i. As published, the central algorithm is formally incomplete.
  4. [§IV-B.1, Table I] Table I mixes different training protocols: ImVoxelNet, NeRF-Det, and CN-RMA are reproduced with 20 training views and 50 test views, while ImGeoNet and NeRF-Det++ are taken from their papers with 50 training views and 50 test views. The comparison is therefore not controlled across all baselines, and the paper's claim of beating 'the second-best approach' should be based on matched protocols. Use Table II's 20-view training setup (which is fair against CN-RMA) as the headline comparison, or retrain ImGeoNet and NeRF-Det++ under the same protocol and report both settings.
  5. [§III-C, Eq. (7), Table VII] Table VII introduces a constant θ that is added to the occupancy scores and reports that θ has a large effect (mAP@0.25 drops from 51.3 to 45.5 when θ=0 on SUN RGB-D). However, Section III-C and Algorithm 1 never define θ or explain where the addition occurs; Eq. (7) only shows V' ← V ⊙ S. This is a reproducibility gap for a component that materially changes performance. Add θ to the method description and state its value in the implementation details.
minor comments (5)
  1. [Table I] In Table I(a), the +4.8 delta is over ImGeoNet while the text reports a 16.9% relative improvement over CN-RMA; clarify which reference each delta uses.
  2. [Figure 1] The Figure 1 caption includes 'Depth supervision (optional)' but the loss in Eq. (8) always includes depth supervision; reconcile the caption with the method.
  3. [§III-B, Eq. (3)] The text uses R_i for the 4x4 extrinsic matrix; use T_i or [R_i | t_i] to avoid confusion with the 3x3 rotation matrix and with the undefined norm in Algorithm 2.
  4. [§IV-B.1] The sentence about data efficiency says results are 'on par' with CN-RMA at 70 views, while Table II shows our 20-view mAP@0.25 is higher (55.6 vs 52.8) and mAP@0.5 is lower (31.5 vs 32.6); phrase this as 'comparable or better depending on the metric.'
  5. [General] The paper does not provide code or model weights; a public implementation would help verify the reported results, especially given the Algorithm 2 notation issue.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the geometric cues are derived from predicted depth, not from detection targets, and all claims are validated against external benchmarks.

full rationale

The paper does not derive detections from first principles; it proposes an architecture and validates it against external benchmark metrics (SUN RGB-D mAP, ScanNetV2 mAP, KITTI AP3D). The two geometry modules are deterministic functions of predicted depth maps. Algorithm 1 computes occupancy scores by back-projecting depth maps to point clouds and counting points per voxel; Algorithm 2 computes a TSDF volume by standard fusion of depth maps. These quantities are combined with the image-derived voxel feature volume via element-wise multiplication (Eq. 7) and concatenation, and the detection head is supervised by bounding-box losses (Eq. 8). The detection targets never define the occupancy or TSDF values, so the reported improvements cannot reduce to the model's inputs by construction. The only tension is the abstract's 'without requiring supervision from 3D signals' wording, which is contradicted by the depth L1 loss lambda * L_depth in Eq. (8) and by the ablation statement that the baseline is 'also trained using ground truth depth maps as supervision' (Sec. IV-C.1); this is a claim-consistency concern, not circularity. The sole self-citation ([39], a CenterNet-style 2D detector) is used only to illustrate existing 2D frameworks and is not load-bearing. The occupancy constant theta tuned on the validation set is ordinary hyperparameter selection rather than a fitted input renamed as a prediction.

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

The method's gains rest on several manually chosen hyperparameters (loss weights, occupancy constant) and on the availability of depth supervision and accurate camera poses. No new physical entities are introduced.

free parameters (6)
  • depth loss weight λ = 0.5
    Chosen via Table VII on SUN RGB-D; affects balance of detection and depth losses.
  • box loss weight α = 2
    Set in Section IV.A.3 for the KITTI outdoor head.
  • direction loss weight β = 0.2
    Set in Section IV.A.3 for the KITTI outdoor head.
  • occupancy score constant θ = 1.0
    Added to occupancy scores in the monocular setting; Table VII shows a large drop without it, and it is tuned on SUN RGB-D.
  • TSDF truncation distance d = unspecified
    An input to Algorithm 2, but its exact value is not reported.
  • 3D volume shape and voxel size = 40x40x16, 0.16m indoor; 216x248x12, 0.32m outdoor
    Resolution choices are fixed by the implementation and are not ablated.
assumptions (3)
  • domain assumption Predicted depth from DPT (Ranftl et al. 2021) is accurate enough to serve as a geometry cue.
    Invoked in Section III-C; if depth is poor, occupancy scores and TSDF values are poor.
  • domain assumption Camera intrinsics and extrinsics are known and accurate.
    Equation (3) and Algorithm 2 use these parameters for back-projection and TSDF fusion; they are treated as given inputs.
  • domain assumption Ground-truth depth maps are available for depth loss supervision.
    Section III-E defines an L1 depth loss and Table V says the baseline is trained using ground-truth depth maps; this contradicts the no-3D-supervision wording.

how reviews work

0 comments
Cite this review

Pith. "Pith review of 3DGeoDet: General-purpose Geometry-aware Image-based 3D Object Detection." pith.science (2026). https://pith.science/paper/MGGXZO3X

@misc{pith2026250609541,
  author       = {Pith},
  title        = {Pith review of: 3DGeoDet: General-purpose Geometry-aware Image-based 3D Object Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MGGXZO3X}},
  note         = {Machine review of arXiv:2506.09541}
}
read the original abstract

This paper proposes 3DGeoDet, a novel geometry-aware 3D object detection approach that effectively handles single- and multi-view RGB images in indoor and outdoor environments, showcasing its general-purpose applicability. The key challenge for image-based 3D object detection tasks is the lack of 3D geometric cues, which leads to ambiguity in establishing correspondences between images and 3D representations. To tackle this problem, 3DGeoDet generates efficient 3D geometric representations in both explicit and implicit manners based on predicted depth information. Specifically, we utilize the predicted depth to learn voxel occupancy and optimize the voxelized 3D feature volume explicitly through the proposed voxel occupancy attention. To further enhance 3D awareness, the feature volume is integrated with an implicit 3D representation, the truncated signed distance function (TSDF). Without requiring supervision from 3D signals, we significantly improve the model's comprehension of 3D geometry by leveraging intermediate 3D representations and achieve end-to-end training. Our approach surpasses the performance of state-of-the-art image-based methods on both single- and multi-view benchmark datasets across diverse environments, achieving a 9.3 mAP@0.5 improvement on the SUN RGB-D dataset, a 3.3 mAP@0.5 improvement on the ScanNetV2 dataset, and a 0.19 AP3D@0.7 improvement on the KITTI dataset. The project page is available at: https://cindy0725.github.io/3DGeoDet/.

Figures

Figures reproduced from arXiv: 2506.09541 by the authors.

Figure 1
Figure 1. Comparison of detection performance and framework with existing [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overall architecture of our 3DGeoDet method. Given either single or multiple posed RGB images, we begin by extracting 2D image features using a transformer backbone. To estimate depth information, we reassemble and fuse multi-scale 2D image features and adopt a depth prediction head. The fused 2D features are back-projected to obtain a 3D feature volume. Since the 3D feature volume lacks explicit geometric informati… view at source ↗
Figure 3
Figure 3. Qualitative results of multi-view 3D object detection on ScanNetV2. We randomly sample 50 input images for each scene during inference. Compared to ImVoxelNet and CN-RMA, our method performs better in predicting objects with smaller sizes or objects in corners, such as the small table, the chairs, and the bathroom door in the red circle. TABLE V EFFECTIVENESS OF VOXEL OCCUPANCY ATTENTION AND TSDF SHAPING. WE CONDUCT… view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Qualitative results of monocular 3D object detection on KITTI. We input one image for each scene. Compared with ImVoxelNet, our method performs better in predicting small, distant, and occluded objects. Ground truth ImVoxelNet Ours bed table sofa chair toilet desk dres…
Figure 5
Figure 5. Figure 5: Qualitative results of monocular 3D object detection on SUN RGB-D. We input one image for each scene. Compared with ImVoxelNet, our method performs better in predicting the rotation angles of large-sized objects such as beds, cabinets, and sofas. TABLE VI IMPACT OF GEN…
Figure 6
Figure 6. Figure 6: Failure cases in SUN RGB-D and ScanNetV2. For the SUN RGB-D dataset, our method predicts tables and chairs that are absent from the ground truth annotations. For the ScanNetV2 dataset, our method misses the small cabinet located in the upper right corner and the curtai…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

64 extracted references · 52 canonical work pages

  1. [1]

    Faster r-cnn: Towards real-time object detection with region proposal networks,

    S. Ren, K. He, R. Girshick, and J. Sun, “Faster r-cnn: Towards real-time object detection with region proposal networks,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 39, no. 6, pp. 1137– 1149, 2017

    2017

  2. [2]

    Open world object detection: A survey,

    Y . Li, Y . Wang, W. Wang, D. Lin, B. Li, and K.-H. Yap, “Open world object detection: A survey,”IEEE Transactions on Circuits and Systems for Video Technology, 2024

    work page 2024

  3. [3]

    Few- shot object detection with multilevel information interaction for optical remote sensing images,

    L. Wang, S. Mei, Y . Wang, J. Lian, Z. Han, and X. Chen, “Few- shot object detection with multilevel information interaction for optical remote sensing images,”IEEE Transactions on Geoscience and Remote Sensing, vol. 62, pp. 1–14, 2024

    work page 2024

  4. [4]

    Neural saliency algorithm guide bi-directional visual perception style transfer,

    C. Zhu, W. Yan, X. Cai, S. Liu, T. H. Li, and G. Li, “Neural saliency algorithm guide bi-directional visual perception style transfer,”CAAI Transactions on Intelligence Technology, vol. 5, no. 1, pp. 1–8, 2020

    work page 2020

  5. [5]

    Pdnet: Prior-model guided depth-enhanced network for salient object detection,

    C. Zhu, X. Cai, K. Huang, T. H. Li, and G. Li, “Pdnet: Prior-model guided depth-enhanced network for salient object detection,” inIEEE International Conference on Multimedia and Expo, 2019, pp. 199–204

    work page 2019

  6. [6]

    Pointrcnn: 3d object proposal generation and detection from point cloud,

    S. Shi, X. Wang, and H. Li, “Pointrcnn: 3d object proposal generation and detection from point cloud,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019, pp. 770– 779

    work page 2019

  7. [7]

    Deep hough voting for 3d object detection in point clouds,

    C. R. Qi, O. Litany, K. He, and L. J. Guibas, “Deep hough voting for 3d object detection in point clouds,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2019, pp. 9277–9286

    work page 2019

  8. [8]

    Std: Sparse-to-dense 3d object detector for point cloud,

    Z. Yang, Y . Sun, S. Liu, X. Shen, and J. Jia, “Std: Sparse-to-dense 3d object detector for point cloud,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2019, pp. 1951–1960

    work page 2019

Show all 64 references
  1. [9]

    From points to parts: 3d object detection from point cloud with part-aware and part-aggregation network,

    S. Shi, Z. Wang, J. Shi, X. Wang, and H. Li, “From points to parts: 3d object detection from point cloud with part-aware and part-aggregation network,”IEEE Transactions on Pattern Analysis and Machine Intelli- gence, vol. 43, no. 8, pp. 2647–2664, 2020

    2020

  2. [10]

    H3dnet: 3d object detection using hybrid geometric primitives,

    Z. Zhang, B. Sun, H. Yang, and Q. Huang, “H3dnet: 3d object detection using hybrid geometric primitives,” inProceedings of the European Conference on Computer Vision, 2020, pp. 311–329

    2020

  3. [11]

    Group-free 3d object detection via transformers,

    Z. Liu, Z. Zhang, Y . Cao, H. Hu, and X. Tong, “Group-free 3d object detection via transformers,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 2949–2958

    2021

  4. [12]

    Fcaf3d: Fully convolu- tional anchor-free 3d object detection,

    D. Rukhovich, A. V orontsova, and A. Konushin, “Fcaf3d: Fully convolu- tional anchor-free 3d object detection,” inProceedings of the European Conference on Computer Vision, 2022, pp. 477–493

    2022

  5. [13]

    Multi-dimensional pruned sparse convolution for efficient 3d object detection,

    L. Li, X. Yue, Z. Xu, and S. Xie, “Multi-dimensional pruned sparse convolution for efficient 3d object detection,” inIEEE International Conference on Image Processing, 2023, pp. 3190–3194

    2023

  6. [14]

    Da-net: Density-aware 3d object detection network for point clouds,

    S. Wang, K. Lu, J. Xue, and Y . Zhao, “Da-net: Density-aware 3d object detection network for point clouds,”IEEE Transactions on Multimedia, pp. 1–14, 2023

    2023

  7. [15]

    Farp-net: Local-global feature aggregation and relation-aware proposals for 3d object detection,

    T. Xie, L. Wang, K. Wang, R. Li, X. Zhang, H. Zhang, L. Yang, H. Liu, and J. Li, “Farp-net: Local-global feature aggregation and relation-aware proposals for 3d object detection,”IEEE Transactions on Multimedia, vol. 26, pp. 1027–1040, 2024

    2024

  8. [16]

    Sp-det: Leveraging saliency prediction for voxel-based 3d ob- ject detection in sparse point cloud,

    P. An, Y . Duan, Y . Huang, J. Ma, Y . Chen, L. Wang, Y . Yang, and Q. Liu, “Sp-det: Leveraging saliency prediction for voxel-based 3d ob- ject detection in sparse point cloud,”IEEE Transactions on Multimedia, vol. 26, pp. 2795–2808, 2024

    2024

  9. [17]

    Imvoxelnet: Image to voxels projection for monocular and multi-view general-purpose 3d object detection,

    D. Rukhovich, A. V orontsova, and A. Konushin, “Imvoxelnet: Image to voxels projection for monocular and multi-view general-purpose 3d object detection,” inProceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2022, pp. 2397–2406

    2022

  10. [18]

    Nerf-det: Learning geometry- aware volumetric representation for multi-view 3d object detection,

    C. Xu, B. Wu, J. Hou, S. Tsai, R. Li, J. Wang, W. Zhan, Z. He, P. Vajda, K. Keutzer, and M. Tomizuka, “Nerf-det: Learning geometry- aware volumetric representation for multi-view 3d object detection,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, ...

    2023

  11. [19]

    Cn-rma: Combined network with ray marching aggregation for 3d indoor object detection from multi- view images,

    G. Shen, J. Huang, Z. Hu, and B. Wang, “Cn-rma: Combined network with ray marching aggregation for 3d indoor object detection from multi- view images,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 21 326–21 335

    2024

  12. [20]

    Scannet: Richly-annotated 3d reconstructions of indoor scenes,

    A. Dai, A. X. Chang, M. Savva, M. Halber, T. Funkhouser, and M. Nießner, “Scannet: Richly-annotated 3d reconstructions of indoor scenes,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2017, pp. 5828–5839. 12

    2017

  13. [21]

    Imgeonet: Image-induced geometry-aware voxel representation for multi-view 3d object detection,

    T. Tu, S.-P. Chuang, Y .-L. Liu, C. Sun, K. Zhang, D. Roy, C.-H. Kuo, and M. Sun, “Imgeonet: Image-induced geometry-aware voxel representation for multi-view 3d object detection,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 6996–7007

    2023

  14. [22]

    Sun rgb-d: A rgb-d scene under- standing benchmark suite,

    S. Song, S. P. Lichtenberg, and J. Xiao, “Sun rgb-d: A rgb-d scene under- standing benchmark suite,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2015, pp. 567–576

    2015

  15. [23]

    Are we ready for autonomous driv- ing? the kitti vision benchmark suite,

    A. Geiger, P. Lenz, and R. Urtasun, “Are we ready for autonomous driv- ing? the kitti vision benchmark suite,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2012

    2012

  16. [24]

    Frustum pointnets for 3d object detection from rgb-d data,

    C. R. Qi, W. Liu, C. Wu, H. Su, and L. J. Guibas, “Frustum pointnets for 3d object detection from rgb-d data,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2018, pp. 918– 927

    2018

  17. [25]

    3dssd: Point-based 3d single stage object detector,

    Z. Yang, Y . Sun, S. Liu, and J. Jia, “3dssd: Point-based 3d single stage object detector,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020, pp. 11 040–11 048

    2020

  18. [26]

    Pointnet: Deep learning on point sets for 3d classification and segmentation,

    C. R. Qi, H. Su, K. Mo, and L. J. Guibas, “Pointnet: Deep learning on point sets for 3d classification and segmentation,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2017, pp. 652–660

    2017

  19. [27]

    Pointnet++: Deep hierarchical feature learning on point sets in a metric space,

    C. R. Qi, L. Yi, H. Su, and L. J. Guibas, “Pointnet++: Deep hierarchical feature learning on point sets in a metric space,”Advances in Neural Information Processing Systems, vol. 30, 2017

    2017

  20. [28]

    Second: Sparsely embedded convolutional detection,

    Y . Yan, Y . Mao, and B. Li, “Second: Sparsely embedded convolutional detection,”Sensors, vol. 18, no. 10, p. 3337, 2018

    2018

  21. [29]

    Pv- rcnn: Point-voxel feature set abstraction for 3d object detection,

    S. Shi, C. Guo, L. Jiang, Z. Wang, J. Shi, X. Wang, and H. Li, “Pv- rcnn: Point-voxel feature set abstraction for 3d object detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020, pp. 10 529–10 538

    2020

  22. [30]

    Ssn: Shape signature networks for multi-class object detection from point clouds,

    X. Zhu, Y . Ma, T. Wang, Y . Xu, J. Shi, and D. Lin, “Ssn: Shape signature networks for multi-class object detection from point clouds,” inProceedings of the European Conference on Computer Vision, 2020, pp. 581–597

    2020

  23. [31]

    Structure aware single-stage 3d object detection from point cloud,

    C. He, H. Zeng, J. Huang, X.-S. Hua, and L. Zhang, “Structure aware single-stage 3d object detection from point cloud,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020, pp. 11 873–11 882

    2020

  24. [32]

    Center-based 3d object detection and tracking,

    T. Yin, X. Zhou, and P. Krahenbuhl, “Center-based 3d object detection and tracking,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 11 784–11 793

    2021

  25. [33]

    Centerformer: Center-based transformer for 3d object detection,

    Z. Zhou, X. Zhao, Y . Wang, P. Wang, and H. Foroosh, “Centerformer: Center-based transformer for 3d object detection,” inProceedings of the European Conference on Computer Vision, 2022, pp. 496–513

    2022

  26. [34]

    Multi-modal fusion based on depth adaptive mechanism for 3d object detection,

    Z. Liu, J. Cheng, J. Fan, S. Lin, Y . Wang, and X. Zhao, “Multi-modal fusion based on depth adaptive mechanism for 3d object detection,” IEEE Transactions on Multimedia, pp. 1–11, 2023

    2023

  27. [35]

    Vpfnet: Improving 3d object detection with virtual point based lidar and stereo data fusion,

    H. Zhu, J. Deng, Y . Zhang, J. Ji, Q. Mao, H. Li, and Y . Zhang, “Vpfnet: Improving 3d object detection with virtual point based lidar and stereo data fusion,”IEEE Transactions on Multimedia, vol. 25, pp. 5291–5304, 2023

    2023

  28. [36]

    Fcos3d: Fully convolutional one- stage monocular 3d object detection,

    T. Wang, X. Zhu, J. Pang, and D. Lin, “Fcos3d: Fully convolutional one- stage monocular 3d object detection,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 913–922

    2021

  29. [37]

    Objects as points,

    X. Zhou, D. Wang, and P. Kr ¨ahenb¨uhl, “Objects as points,” inArXiv Preprint ArXiv:1904.07850, 2019

    1904 arXiv

  30. [38]

    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 and Machine Intelligence, vol. 44, no. 4, pp. 1922–1933, 2020

    1922

  31. [39]

    A self-training approach for point-supervised object detection and counting in crowds,

    Y . Wang, J. Hou, X. Hou, and L.-P. Chau, “A self-training approach for point-supervised object detection and counting in crowds,”IEEE Transactions on Image Processing, vol. 30, pp. 2876–2887, 2021

    2021

  32. [40]

    Pseudo-lidar from visual depth estimation: Bridging the gap in 3d object detection for autonomous driving,

    Y . Wang, W.-L. Chao, D. Garg, B. Hariharan, M. Campbell, and K. Q. Weinberger, “Pseudo-lidar from visual depth estimation: Bridging the gap in 3d object detection for autonomous driving,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 201...

    2019

  33. [41]

    Accurate monocular 3d object detection via color-embedded 3d reconstruction for autonomous driving,

    X. Ma, Z. Wang, H. Li, P. Zhang, W. Ouyang, and X. Fan, “Accurate monocular 3d object detection via color-embedded 3d reconstruction for autonomous driving,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2019, pp. 6851–6860

    2019

  34. [42]

    Monocular 3d object detection with pseudo- lidar point cloud,

    X. Weng and K. Kitani, “Monocular 3d object detection with pseudo- lidar point cloud,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2019, pp. 857–866

    2019

  35. [43]

    Detr3d: 3d object detection from multi-view images via 3d-to-2d queries,

    Y . Wang, V . C. Guizilini, T. Zhang, Y . Wang, H. Zhao, and J. Solomon, “Detr3d: 3d object detection from multi-view images via 3d-to-2d queries,” inConference on Robot Learning, 2022, pp. 180–191

    2022

  36. [44]

    Bevdet: High- performance multi-camera 3d object detection in bird-eye-view,

    J. Huang, G. Huang, Z. Zhu, Y . Ye, and D. Du, “Bevdet: High- performance multi-camera 3d object detection in bird-eye-view,”ArXiv Preprint ArXiv:2112.11790, 2021

    2021 arXiv

  37. [45]

    Petr: Position embedding transformation for multi-view 3d object detection,

    Y . Liu, T. Wang, X. Zhang, and J. Sun, “Petr: Position embedding transformation for multi-view 3d object detection,” inProceedings of the European Conference on Computer Vision, 2022, pp. 531–548

    2022

  38. [46]

    Monodetr: Depth-guided transformer for monocular 3d object detection,

    R. Zhang, H. Qiu, T. Wang, Z. Guo, Z. Cui, Y . Qiao, H. Li, and P. Gao, “Monodetr: Depth-guided transformer for monocular 3d object detection,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 9155–9166

    2023

  39. [47]

    Crossdtr: Cross-view and depth-guided transformers for 3d object detection,

    C.-Y . Tseng, Y .-R. Chen, H.-Y . Lee, T.-H. Wu, W.-C. Chen, and W. H. Hsu, “Crossdtr: Cross-view and depth-guided transformers for 3d object detection,” in2023 IEEE International Conference on Robotics and Automation, 2023, pp. 4850–4857

    2023

  40. [48]

    Petrv2: A unified framework for 3d perception from multi-camera images,

    Y . Liu, J. Yan, F. Jia, S. Li, A. Gao, T. Wang, and X. Zhang, “Petrv2: A unified framework for 3d perception from multi-camera images,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 3262–3272

    2023

  41. [49]

    Monodtr: Monocular 3d object detection with depth-aware transformer,

    K.-C. Huang, T.-H. Wu, H.-T. Su, and W. H. Hsu, “Monodtr: Monocular 3d object detection with depth-aware transformer,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 4012–4021

    2022

  42. [50]

    Occformer: Dual-path transformer for vision-based 3d semantic occupancy prediction,

    Y . Zhang, Z. Zhu, and D. Du, “Occformer: Dual-path transformer for vision-based 3d semantic occupancy prediction,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 9433–9443

    2023

  43. [51]

    Occ3d: A large-scale 3d occupancy prediction benchmark for autonomous driving,

    X. Tian, T. Jiang, L. Yun, Y . Mao, H. Yang, Y . Wang, Y . Wang, and H. Zhao, “Occ3d: A large-scale 3d occupancy prediction benchmark for autonomous driving,”Advances in Neural Information Processing Systems, vol. 36, 2024

    2024

  44. [52]

    A volumetric method for building complex models from range images,

    B. Curless and M. Levoy, “A volumetric method for building complex models from range images,” inProceedings of the 23rd Annual Con- ference on Computer Graphics and Interactive Techniques, 1996, pp. 303–312

    1996

  45. [53]

    Atlas: End-to-end 3d scene reconstruction from posed images,

    Z. Murez, T. Van As, J. Bartolozzi, A. Sinha, V . Badrinarayanan, and A. Rabinovich, “Atlas: End-to-end 3d scene reconstruction from posed images,” inProceedings of the European Conference on Computer Vision, 2020, pp. 414–431

    2020

  46. [54]

    Neuralrecon: Real-time coherent 3d reconstruction from monocular video,

    J. Sun, Y . Xie, L. Chen, X. Zhou, and H. Bao, “Neuralrecon: Real-time coherent 3d reconstruction from monocular video,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 15 598–15 607

    2021

  47. [55]

    Neural 3d scene reconstruction with the manhattan-world assumption,

    H. Guo, S. Peng, H. Lin, Q. Wang, G. Zhang, H. Bao, and X. Zhou, “Neural 3d scene reconstruction with the manhattan-world assumption,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 5511–5520

    2022

  48. [56]

    V olumefusion: Deep depth fusion for 3d scene reconstruction,

    J. Choe, S. Im, F. Rameau, M. Kang, and I. S. Kweon, “V olumefusion: Deep depth fusion for 3d scene reconstruction,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 16 086–16 095

    2021

  49. [57]

    Dinov2: Learning robust visual features without supervision,

    M. Oquab, T. Darcet, T. Moutakanni, H. V o, M. Szafraniec, V . Khalidov, P. Fernandez, D. Haziza, F. Massa, A. El-Noubyet al., “Dinov2: Learning robust visual features without supervision,”ArXiv Preprint ArXiv:2304.07193, 2023

    2023 arXiv

  50. [58]

    Vision transformers for dense prediction,

    R. Ranftl, A. Bochkovskiy, and V . Koltun, “Vision transformers for dense prediction,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 12 179–12 188

    2021

  51. [59]

    Mononerd: Nerf-like representations for monocular 3d object detection,

    J. Xu, L. Peng, H. Cheng, H. Li, W. Qian, K. Li, W. Wang, and D. Cai, “Mononerd: Nerf-like representations for monocular 3d object detection,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 6814–6824

    2023

  52. [60]

    Nerf-det++: Incorporating semantic cues and perspective- aware depth supervision for indoor multi-view 3d detection,

    C. Huang, Y . Hou, W. Ye, D. Huang, X. Huang, B. Lin, D. Cai, and W. Ouyang, “Nerf-det++: Incorporating semantic cues and perspective- aware depth supervision for indoor multi-view 3d detection,”ArXiv Preprint ArXiv:2402.14464, 2024

    2024 arXiv

  53. [61]

    Did-m3d: Decoupling instance depth for monocular 3d object detection,

    L. Peng, X. Wu, Z. Yang, H. Liu, and D. Cai, “Did-m3d: Decoupling instance depth for monocular 3d object detection,” inProceedings of the European Conference on Computer Vision, 2022

    2022

  54. [62]

    Monouni: A unified vehicle and infrastructure- side monocular 3d object detection network with sufficient depth clues,

    J. Jia, Z. Li, and Y . Shi, “Monouni: A unified vehicle and infrastructure- side monocular 3d object detection network with sufficient depth clues,” inAdvances in Neural Information Processing Systems, vol. 36, 2023

    2023

  55. [63]

    Monolss: Learnable sample selection for monocular 3d detection,

    Z. Li, J. Jia, and Y . Shi, “Monolss: Learnable sample selection for monocular 3d detection,” inInternational Conference on 3D Vision, 2024, pp. 1125–1135

    2024

  56. [64]

    MMDetection3D: OpenMMLab next-generation plat- form for general 3d object detection,

    M. Contributors, “MMDetection3D: OpenMMLab next-generation plat- form for general 3d object detection,” https://github.com/open-mmlab/ mmdetection3d, 2020. 13 Yi Zhangreceived her B.Sc. degree in Computer Science from the Hong Kong University of Science and Technology in 2020 ...

    2020

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.