Pith. sign in

REVIEW 2 major objections 5 minor 1 cited by

OC-SOP: Enhancing Vision-Based 3D Semantic Occupancy Prediction by Object-Centric Awareness

T0 review · 2 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Pith's one-line read Detected object boxes sharpen camera-only 3D occupancy prediction

desk verdict OC-SOP's object-centric fusion idea is worth a look, but the paper's headline numbers are internally inconsistent and the SOTA claim is not currently verifiable. read the letter →

arxiv 2506.18798 v2 pith:LSKU7XIX submitted 2025-06-23 cs.CV cs.AIcs.RO

classification cs.CVcs.AIcs.RO
keywords semanticoccupancypredictionscenecompletionobject-centricawareness3Dobjectdetectiondeformablecross-attentioncamera-basedperceptionautonomousdrivingmulti-tasklearning
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 camera-based 3D semantic occupancy prediction improves when the network is made explicitly aware of objects, not just voxels. It adds a detection branch that turns predicted 3D bounding boxes into queries and fuses them into a semantic completion U-Net with deformable cross-attention. On SemanticKITTI, this raises mean intersection-over-union from 12.20 to 14.83 and, for the eight foreground classes, from 4.24 to 7.95, which the authors report as state of the art. Sympathetic readers should care because dynamic foreground objects are exactly the errors that matter most for safe driving.

What carries the argument

The load-bearing mechanism is box-feature fusion through deformable cross-attention. Each surviving detection proposal is tokenized by an MLP into a query $q$; the query samples $K$ learned offset locations around a base position $p$ in the completion U-Net's latent space and aggregates the values there with attention weights, $\text{Output}(q)=\sum_{k=1}^{K} A_k\, V(p+\Delta p_k)$. This lets object hypotheses read and inject information at the voxel features where they sit, rather than being concatenated uniformly across the whole volume. A three-stage training scheme, which first warms up the main branch, then trains the detection head, and finally fine-tunes jointly with the detection head frozen, is what makes the two-branch system stable enough to deliver the reported results.

What would settle it

A decisive check is to measure class-conditional detection recall on SemanticKITTI validation: report the fraction of ground-truth objects of each of the eight foreground classes whose centers fall within 1 m of an NMS-selected proposal; if recall is near zero for bicycle, motorcycle, and truck and yet the foreground mIoU gain from 3.41 to 7.95 persists, the improvement cannot be attributed to the object-specific content of the fused queries.

Watch

Extended reading notes

Core claim

OC-SOP's central claim is that high-level object-centric cues, namely box center, size, orientation, class, and objectness, can act as constraints that keep voxel predictions from merging neighboring objects, truncating shapes, or hallucinating vehicles. The paper demonstrates this by comparing its full two-branch model with its own main-branch-only ablation on the SemanticKITTI validation split: foreground mIoU rises from 3.41 to 7.95 when detection cues are fused, while background mIoU rises only from 18.81 to 19.83. On the test set, classes that other methods mostly miss, such as bicycle at 7.80 IoU, motorcycle at 6.30, and person at 5.80, become recognizable, and overall mIoU reaches 14.83.

Load-bearing premise

The method assumes that boxes from a detector trained only on Car, Pedestrian, and Cyclist transfer usefully to eight finer SemanticKITTI classes, including classes the detector has never seen.

Editorial extensions

If this is right

  • SemanticKITTI test performance reaches 14.83 mIoU and 43.30 IoU, exceeding the 12.20 mIoU of the strongest compared vision baseline.
  • Foreground-object mIoU improves from 4.24 to 7.95, with the largest relative gains in rare classes such as bicycle, motorcycle, and person.
  • On the validation split, adding object-centric fusion is worth about 4.54 percentage points of foreground mIoU and about 1.02 percentage points of background mIoU.
  • Even without the detection branch, the encoder-dual-decoder backbone beats the MonoScene baseline, so disentangling depth and semantic feature extraction helps occupancy prediction independently.
  • A detector trained on KITTI's three coarse classes can transfer enough semantic prior to serve the eight finer SemanticKITTI foreground classes.

Reading between the lines

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

  • Because only three coarse classes supervise the detector, the method's ceiling is likely set by how well those priors cover rare SemanticKITTI classes; a natural next test is to train the detection branch on more granular labels and measure whether the remaining foreground gap closes.
  • The same query-fusion mechanism could be applied to LiDAR-camera fusion, where box proposals are easier to obtain and might give even stronger boundary constraints for occupancy completion.
  • Replacing the supervised detector with a class-agnostic proposal source, such as objectness from depth or motion cues, would separate whether object-centric structure itself helps from whether semantic detection supervision is required.
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

2 major / 5 minor

Summary. The paper proposes OC-SOP, a camera-based 3D semantic occupancy prediction model with a dual-branch architecture: a main completion branch (encoder with dual decoders, depth-aware feature lifting, and a 3D completion U-Net) and an object detection branch that produces box proposals, tokenizes them into queries, and fuses them into the completion branch via deformable cross-attention. Training is staged: warm-up of the main branch, detection-head training on KITTI 3D object detection data, then joint fine-tuning with the detection head frozen. The authors report state-of-the-art SemanticKITTI test-set results (14.83 mIoU, 43.30 IoU, foreground mIoU 7.95) and an ablation study on the validation set that attributes a large foreground-object gain to the object-centric fusion module.

Significance. If the reported results were reliable, the paper would make a useful contribution: it is among the first to inject explicit object-centric box cues into monocular semantic scene completion, and the clean ablation (Settings I vs. III) isolates a 4.54-point foreground mIoU improvement attributable to the object-centric fusion. The architectural idea is well motivated and the writing is generally clear. However, the quantitative evidence as presented contains a load-bearing inconsistency that currently prevents verification of the state-of-the-art claim. The reported ablation design is a genuine strength, but the tables must be corrected and verified before the contribution can be assessed.

major comments (2)
  1. [Table I and Table II] The central experimental evidence is internally inconsistent. The OC-SOP row in Table I (labeled SemanticKITTI test set) and Setting I in Table II (labeled SemanticKITTI validation set) report exactly the same per-class IoU for all 19 classes and the same overall IoU of 43.30, even though the SemanticKITTI test and validation splits are disjoint; identical per-class values to one decimal place across disjoint splits is effectively impossible, so at least one table is mislabeled. In addition, the mean of the 19 per-class values in Table II Setting I is 14.83, not the printed mIoU of 14.56; rounding cannot explain a 0.27 gap over nineteen classes. Because the paper's headline claims (state-of-the-art 14.83 mIoU on the test set and 7.95 foreground mIoU) rest entirely on these numbers, the authors must correct the split labels, reconcile the mIoU arithmetic, and provide the official SemanticKITTI test-server evaluations or equivalent verifiable evidence before the claims can be assessed.
  2. [Section IV.A.1 / Section III.B] The object-centric mechanism is trained on the KITTI 3D object detection dataset with only three classes (Car, Pedestrian, Cyclist), yet Tables I and II report per-class gains for eight SemanticKITTI foreground classes, including truck, motorcycle, other-vehicle, person, bicyclist, and motorcyclist. The paper provides no mapping or mechanism by which the detection branch's three-class output produces class-specific cues for these eight classes; the statement that 'these rough labels still provide a semantic prior' (Section IV.A.1) does not explain how the fusion module can distinguish, for example, bicyclist from motorcyclist. The authors should clarify whether the detection branch's class output is used at all during fusion and, if so, specify the KITTI-to-SemanticKITTI label mapping. Without this, the attribution of the per-class foreground gains to object-centric awareness is not supported.
minor comments (5)
  1. [Section III.C, Eq. (2)] Please define the notation in Equation (2): the base position p, the learned offset Delta-p_k, the value function V(.), and the range or normalization of the attention weights A_k are not specified, which makes the deformable attention description incomplete.
  2. [Section III.A, reference [32]] The citation [32] for the 'dual-branch network architecture' points to a pan-sharpening paper (DUDB) that appears unrelated to the described occupancy architecture; please replace it with a relevant reference or remove the citation.
  3. [Section IV.A.1] The sentence claiming that KITTI's three annotated foreground categories 'completely align with our definition of foreground objects' is stronger than the immediately following caveat about coarser labels; please rephrase to acknowledge the coarse-to-fine transfer issue that the paper itself raises.
  4. [Figure 4 and Section IV.B] There are minor typographical errors, including 'V oxels' in the Figure 4 caption and inconsistent use of 'VoxFormer' in the text and tables.
  5. [Table I / Section IV.B] The paper quotes foreground mIoU values of 7.95 for OC-SOP and 4.24 for VoxFormer but does not report these aggregates in the table; please include the foreground mIoU as an explicit row in Table I so readers can verify the claimed 3.71-point gain.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the object-centric gain is measured on held-out SemanticKITTI, with detection cues trained on the separate KITTI benchmark, so the reported improvement does not reduce to a fitted input or self-citation.

full rationale

The paper's central claim is an architectural contribution: object detection cues extracted by a detection branch are fused into a 3D occupancy completion network. The detection branch is trained on the KITTI 3D object detection benchmark, with its own coarse labels, while the occupancy and semantic supervision comes from SemanticKITTI. The foreground-object gain is therefore not a fitted-input prediction: the box proposals are not constructed from the target voxel labels, and the fusion module is trained on held-out splits. The ablations compare fusion settings on the validation set, which is a controlled comparison rather than a tautology. No uniqueness theorem from the authors is invoked to force a choice, and the self-citations in the reference list (SWA-SOP, DiffSSC, SLCF-Net) are related prior work, not load-bearing inputs to the reported derivation. No equation defines an output as an input, and no fitted parameter is renamed as a prediction. The skeptic's concerns about Table I and Table II—identical per-class values despite nominally different splits, and a printed mIoU (14.56) that does not match the arithmetic mean of its own per-class rows—are evidence-quality and internal-consistency issues, not circularity. Under the stated rules, no circular step can be exhibited, so the score is 0.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The central claim rests on the validity of the SemanticKITTI and KITTI ground truth, on the transferability of coarse detection labels to fine-grained occupancy classes, and on the stability of the three-stage training schedule. No new physical or mathematical entities are introduced; the free parameters listed are hand-chosen hyperparameters that shape the detection branch and fusion module.

free parameters (6)
  • objectness filter threshold = 0.2
    Used to filter low-score detection proposals during joint fine-tuning and inference, directly controlling which boxes become fusion queries.
  • NMS IoU threshold = 0.7
    Non-maximum suppression threshold over object proposals, determining the final set of box features passed to the deformable cross-attention fusion.
  • positive/negative center assignment margins = 1 m / 2 m
    Candidate detection centers within 1 m of a ground-truth object center are positive, farther than 2 m are negative; this assignment shapes the detection loss.
  • stage-wise training epochs = 5 / 10 / 10
    Warm-up, detection-only, and joint fine-tuning epochs; the schedule is hand-set and affects the stability and final performance.
  • deformable attention sampling points K = not specified
    Number of sampling points per box query in the fusion module is not reported, yet it is a hyperparameter of Eq. 2.
  • box regression loss component weights = not specified
    Different weights for center, size, and orientation in the Smooth L1 regression loss are mentioned but their values are omitted.
assumptions (5)
  • domain assumption SemanticKITTI ground truth voxelization and class labels are accurate for training and evaluation.
    The model is trained and evaluated on SemanticKITTI's aggregated LiDAR scans. The paper itself notes that dynamic objects may exhibit ghosting artifacts in the ground truth (Sec. IV.B), which could affect the reported metrics.
  • domain assumption KITTI 3D detection labels provide a valid semantic prior for SemanticKITTI foreground classes.
    The detection branch is trained only on KITTI classes (Car, Pedestrian, Cyclist) but produces cues for 8 SemanticKITTI foreground classes. The paper explicitly relies on this transfer, which is the weakest assumption in the evaluation.
  • domain assumption The encoder dual decoder (EDD) depth estimates are accurate enough for feature lifting.
    The main branch lifts 2D semantic features into 3D using the predicted depth distribution. Incorrect depth would misplace features and degrade both occupancy and detection quality.
  • standard math The deformable cross-attention fusion (Eq. 2) is implemented faithfully and converges.
    The fusion module follows the standard deformable attention formulation with learned offsets; the paper provides no convergence guarantees, only empirical results.
  • domain assumption The three-stage training strategy does not cause catastrophic forgetting or instability.
    Stage 2 trains the detection head without the completion U-Net, and Stage 3 freezes the head while jointly fine-tuning. The success depends on the schedule remaining stable, which is asserted but not analyzed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of OC-SOP: Enhancing Vision-Based 3D Semantic Occupancy Prediction by Object-Centric Awareness." pith.science (2026). https://pith.science/paper/LSKU7XIX

@misc{pith2026250618798,
  author       = {Pith},
  title        = {Pith review of: OC-SOP: Enhancing Vision-Based 3D Semantic Occupancy Prediction by Object-Centric Awareness},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LSKU7XIX}},
  note         = {Machine review of arXiv:2506.18798}
}
read the original abstract

Autonomous driving perception faces significant challenges due to occlusions and incomplete scene data in the environment. To overcome these issues, the task of semantic occupancy prediction (SOP) is proposed, which aims to jointly infer both the geometry and semantic labels of a scene from images. However, conventional camera-based methods typically treat all categories equally and primarily rely on local features, leading to suboptimal predictions, especially for dynamic foreground objects. To address this, we propose Object-Centric SOP (OC-SOP), a framework that integrates high-level object-centric cues extracted via a detection branch into the semantic occupancy prediction pipeline. This object-centric integration significantly enhances the prediction accuracy for foreground objects and achieves state-of-the-art performance among all categories on SemanticKITTI.

Figures

Figures reproduced from arXiv: 2506.18798 by the authors.

Figure 1
Figure 1. OC-SOP predicts a semantic occupancy volume from a [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The overall pipeline of OC-SOP can be divided into a main prediction branch (top) and an object detection-based branch (bottom). [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Three-stage training strategy of OC-SOP. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Qualitative results on the SemanticKITTI validation set. Only the 19 semantic classes are visualized; voxels labeled as empty [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. SWA-SOP: Spatially-aware Window Attention for Semantic Occupancy Prediction in Autonomous Driving

    cs.CV 2025-06 conditional novelty 6.0 of 10

    SWA-SOP shows that sliding-window attention with per-slot spatial embeddings and a center query improves semantic occupancy prediction on LiDAR and camera inputs, but the headline benchmark numbers are second-best rat...

Reference graph

Works this paper leans on

42 extracted references · 33 canonical work pages · cited by 1 Pith paper

  1. [1]

    Anisotropic convolutional networks for 3D semantic scene completion,

    J. Li, K. Han, P. Wang, Y . Liu, and X. Yuan, “Anisotropic convolutional networks for 3D semantic scene completion,” in IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , 2020, pp. 3351– 3359

  2. [2]

    MonoScene: Monocular 3D semantic scene completion,

    A.-Q. Cao and R. de Charette, “MonoScene: Monocular 3D semantic scene completion,” in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 3991–4001

  3. [3]

    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,” IEEE International Conference on Computer Vision (ICCV) , 2023

  4. [4]

    Tri-perspective view for vision-based 3D semantic occupancy prediction,

    Y . Huang, W. Zheng, Y . Zhang, J. Zhou, and J. Lu, “Tri-perspective view for vision-based 3D semantic occupancy prediction,” IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , 2023

  5. [5]

    Multi- view stereo for community photo collections,

    M. Goesele, N. Snavely, B. Curless, H. Hoppe, and S. M. Seitz, “Multi- view stereo for community photo collections,” in IEEE International Conference on Computer Vision (ICCV) , 2007, pp. 1–8

  6. [6]

    Accurate, dense, and robust multiview stereopsis,

    Y . Furukawa and J. Ponce, “Accurate, dense, and robust multiview stereopsis,” IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), vol. 32, no. 8, pp. 1362–1376, 2009

  7. [7]

    ICG-MVSNet: Learning Intra-view and Cross-view Relationships for Guidance in Multi-View Stereo

    Y . Hu, J. Zhang, Z. Zhang, R. Weilharter, Y . Rao, K. Chen, R. Yuan, and F. Fraundorfer, “ICG-MVSNet: Learning intra-view and cross- view relationships for guidance in multi-view stereo,” arXiv preprint arXiv:2503.21525, 2025

  8. [8]

    3D map reconstruction using a monocular camera for smart cities,

    Y . Hu, T. Fu, G. Niu, Z. Liu, and M.-O. Pun, “3D map reconstruction using a monocular camera for smart cities,” Journal of Supercomput- ing, 2022

Show all 42 references
  1. [9]

    LSD-SLAM: Large-scale direct monocular SLAM,

    J. Engel, T. Sch ¨ops, and D. Cremers, “LSD-SLAM: Large-scale direct monocular SLAM,” in European Conference on Computer Vision (ECCV), 2014, pp. 834–849

  2. [10]

    ORB-SLAM2: An open-source SLAM system for monocular, stereo, and RGB-D cameras,

    R. Mur-Artal and J. D. Tard ´os, “ORB-SLAM2: An open-source SLAM system for monocular, stereo, and RGB-D cameras,” IEEE Transactions on Robotics , vol. 33, no. 5, pp. 1255–1262, 2017

  3. [11]

    Depth map prediction from a single image using a multi-scale deep network,

    D. Eigen, C. Puhrsch, and R. Fergus, “Depth map prediction from a single image using a multi-scale deep network,” Advances in Neural Information Processing Systems (NeurIPS) , vol. 27, 2014

  4. [12]

    Learning a probabilistic latent space of object shapes via 3D generative- adversarial modeling,

    J. Wu, C. Zhang, T. Xue, B. Freeman, and J. Tenenbaum, “Learning a probabilistic latent space of object shapes via 3D generative- adversarial modeling,” Advances in Neural Information Processing Systems (NeurIPS), vol. 29, 2016

  5. [13]

    A point set generation network for 3D object reconstruction from a single image,

    H. Fan, H. Su, and L. J. Guibas, “A point set generation network for 3D object reconstruction from a single image,” in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017, pp. 605–613

  6. [14]

    Occupancy networks: Learning 3D reconstruction in function space,

    L. Mescheder, M. Oechsle, M. Niemeyer, S. Nowozin, and A. Geiger, “Occupancy networks: Learning 3D reconstruction in function space,” in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2019, pp. 4460–4470

  7. [15]

    Convolutional occupancy networks,

    S. Peng, M. Niemeyer, L. Mescheder, M. Pollefeys, and A. Geiger, “Convolutional occupancy networks,” in European Conference on Computer Vision (ECCV) , 2020, pp. 523–540

  8. [16]

    Panoptic 3D scene reconstruction from a single RGB image,

    M. Dahnert, J. Hou, M. Nießner, and A. Dai, “Panoptic 3D scene reconstruction from a single RGB image,” Advances in Neural Infor- mation Processing Systems (NeurIPS) , vol. 34, pp. 8282–8293, 2021

  9. [17]

    α-OCC: Uncertainty-aware camera-based 3D semantic scene completion,

    S. Su, N. Chen, F. Juefei-Xu, C. Feng, and F. Miao, “ α-OCC: Uncertainty-aware camera-based 3D semantic scene completion,” arXiv preprint arXiv:2406.11021 , 2024

  10. [18]

    LMSCNet: Lightweight multiscale 3D semantic completion,

    L. Roldao, R. de Charette, and A. Verroust-Blondet, “LMSCNet: Lightweight multiscale 3D semantic completion,” in International Conference on 3D Vision (3DV) , 2020, pp. 111–119

  11. [19]

    Sparse single sweep LiDAR point cloud segmentation via learning contextual shape priors from scene completion,

    X. Yan, J. Gao, J. Li, R. Zhang, Z. Li, R. Huang, and S. Cui, “Sparse single sweep LiDAR point cloud segmentation via learning contextual shape priors from scene completion,” in National Conference on Artificial Intelligence (AAAI) , 2021, pp. 3101–3109

  12. [20]

    SW A-SOP: Spatially-aware window attention for semantic occupancy prediction in autonomous driving,

    H. Cao, R. Materla, and S. Behnke, “SW A-SOP: Spatially-aware window attention for semantic occupancy prediction in autonomous driving,” arXiv preprint arXiv:2506.18785 , 2025

  13. [21]

    SLCF-Net: Sequential LiDAR-camera fusion for semantic scene completion using a 3D recurrent U-Net,

    H. Cao and S. Behnke, “SLCF-Net: Sequential LiDAR-camera fusion for semantic scene completion using a 3D recurrent U-Net,” in IEEE International Conference on Robotics and Automation (ICRA) , 2024, pp. 2767–2773

  14. [22]

    3D semantic scene completion: A survey,

    L. Rold ˜ao, R. de Charette, and A. Verroust-Blondet, “3D semantic scene completion: A survey,”International Journal of Computer Vision (IJCV), vol. 130, no. 8, pp. 1978–2005, 2022

  15. [23]

    Semantic scene completion from a single depth image,

    S. Song, F. Yu, A. Zeng, A. X. Chang, M. Savva, and T. Funkhouser, “Semantic scene completion from a single depth image,” in IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , 2017, pp. 1746–1754

  16. [24]

    DiffSSC: Semantic LiDAR scan comple- tion using denoising diffusion probabilistic models,

    H. Cao and S. Behnke, “DiffSSC: Semantic LiDAR scan comple- tion using denoising diffusion probabilistic models,” arXiv preprint arXiv:2409.18092, 2024

  17. [25]

    ScanComplete: Large-scale scene completion and semantic segmen- tation for 3D scans,

    A. Dai, D. Ritchie, M. Bokeloh, S. Reed, J. Sturm, and M. Nießner, “ScanComplete: Large-scale scene completion and semantic segmen- tation for 3D scans,” in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018, pp. 4578–4587

  18. [26]

    Semantically-aware neural radiance fields for vi- sual scene understanding: A comprehensive review,

    T.-A.-Q. Nguyen, A. Bourki, M. Macudzinski, A. Brunel, and M. Bennamoun, “Semantically-aware neural radiance fields for vi- sual scene understanding: A comprehensive review,” arXiv preprint arXiv:2402.11141, 2024

  19. [27]

    V oxFormer: Sparse voxel transformer for camera- based 3D semantic scene completion,

    Y . Li, Z. Yu, C. Choy, C. Xiao, J. M. Alvarez, S. Fidler, C. Feng, and A. Anandkumar, “V oxFormer: Sparse voxel transformer for camera- based 3D semantic scene completion,” in IEEE Conference on Com- puter Vision and Pattern Recognition (CVPR) , 2023, pp. 9087–9098

  20. [28]

    Adaptive fusion of lidar features for 3d object detection in autonomous driving,

    M. Wang, D. Li, J. R. Casas, and J. Ruiz-Hidalgo, “Adaptive fusion of lidar features for 3d object detection in autonomous driving,” Sensors, 2025

  21. [29]

    End-to-end object detection with transformers,

    N. Carion, F. Massa, G. Synnaeve, N. Usunier, A. Kirillov, and S. Zagoruyko, “End-to-end object detection with transformers,” in European Conference on Computer Vision (ECCV) , 2020, pp. 213– 229

  22. [30]

    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,” in Proceedings of Machine Learning Research (PMLR) , 2022, pp. 180–191

  23. [31]

    Object-centric video prediction via decoupling of object dynamics and interactions,

    A. Villar-Corrales, I. Wahdan, and S. Behnke, “Object-centric video prediction via decoupling of object dynamics and interactions,” in IEEE International Conference on Image Processing (ICIP) , 2023, pp. 1234–1238

  24. [32]

    DUDB: Deep unfolding- based Dual-Branch feature fusion network for pan-sharpening remote sensing images,

    H. Tao, J. Li, Z. Hua, and F. Zhang, “DUDB: Deep unfolding- based Dual-Branch feature fusion network for pan-sharpening remote sensing images,” 2023

  25. [33]

    Deep Hough voting for 3D object detection in point clouds,

    C. R. Qi, W. Liu, C. Wu, H. Su, and L. J. Guibas, “Deep Hough voting for 3D object detection in point clouds,” in IEEE International Conference on Computer Vision (ICCV) , 2019

  26. [34]

    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,” in IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , 2019

  27. [35]

    SemanticKITTI: A dataset for semantic scene un- derstanding of LiDAR sequences,

    J. Behley, M. Garbade, A. Milioto, J. Quenzel, S. Behnke, C. Stach- niss, and J. Gall, “SemanticKITTI: A dataset for semantic scene un- derstanding of LiDAR sequences,” in IEEE International Conference on Computer Vision (ICCV) , 2019

  28. [36]

    Are we ready for autonomous driving? The KITTI vision benchmark suite,

    A. Geiger, P. Lenz, and R. Urtasun, “Are we ready for autonomous driving? The KITTI vision benchmark suite,” in IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , 2012, pp. 3354– 3361

  29. [37]

    Monocular 3D object detection for autonomous driving,

    X. Chen, K. Kundu, Z. Zhang, H. Ma, S. Fidler, and R. Urtasun, “Monocular 3D object detection for autonomous driving,” in IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , 2016, pp. 2147–2156

  30. [38]

    Source-free domain adaptation with class prototype discovery,

    L. Zhou, N. Li, M. Ye, X. Zhu, and S. Tang, “Source-free domain adaptation with class prototype discovery,” Pattern Recognition, 2024

  31. [39]

    Self- Prompting analogical reasoning for UA V object detection,

    N. Li, M. Ye, L. Zhou, S. Tang, Y . Gan, Z. Liang, and X. Zhu, “Self- Prompting analogical reasoning for UA V object detection,” inNational Conference on Artificial Intelligence (AAAI) , 2025

  32. [40]

    Prompt categories cluster for weakly supervised semantic segmenta- tion,

    W. Wu, X. Qiu, S. Song, Z. Chen, X. Huang, F. Ma, and J. Xiao, “Prompt categories cluster for weakly supervised semantic segmenta- tion,” in IEEE Conference on Computer Vision and Pattern Recogni- tion (CVPR), 2025, pp. 3198–3207

  33. [41]

    LLM-Enhanced multimodal fusion for Cross-Domain sequential rec- ommendation,

    W. Wu, Z. Chen, X. Qiu, S. Song, X. Huang, F. Ma, and J. Xiao, “LLM-Enhanced multimodal fusion for Cross-Domain sequential rec- ommendation,” arXiv preprint arXiv:2506.17966 , 2025

  34. [42]

    Image fusion for cross-domain sequential recommendation,

    W. Wu, S. Song, X. Qiu, X. Huang, F. Ma, and J. Xiao, “Image fusion for cross-domain sequential recommendation,” 2025

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.