ZoomNet: Part-Aware Adaptive Zooming Neural Network for 3D Object Detection
Open accessAbstract
3D object detection is an essential task in autonomous driving and robotics. Though great progress has been made, challenges remain in estimating 3D pose for distant and occluded objects. In this paper, we present a novel framework named ZoomNet for stereo imagery-based 3D detection. The pipeline of ZoomNet begins with an ordinary 2D object detection model which is used to obtain pairs of left-right bounding boxes. To further exploit the abundant texture cues in rgb images for more accurate disparity estimation, we introduce a conceptually straight-forward module – adaptive zooming, which simultaneously resizes 2D instance bounding boxes to a unified resolution and adjusts the camera intrinsic parameters accordingly. In this way, we are able to estimate higher-quality disparity maps from the resized box images then construct dense point clouds for both nearby and distant objects. Moreover, we introduce to learn part locations as complementary features to improve the resistance against occlusion and put forward the 3D fitting score to better estimate the 3D detection quality. Extensive experiments on the popular KITTI 3D detection dataset indicate ZoomNet surpasses all previous state-of-the-art methods by large margins (improved by 9.4% on APbv (IoU=0.7) over pseudo-LiDAR). Ablation study also demonstrates that our adaptive zooming strategy brings an improvement of over 10% on AP3d (IoU=0.7). In addition, since the official KITTI benchmark lacks fine-grained annotations like pixel-wise part locations, we also present our KFG dataset by augmenting KITTI with detailed instance-wise annotations including pixel-wise part location, pixel-wise disparity, etc.. Both the KFG dataset and our codes will be publicly available at https://github.com/detectRecog/ZoomNet.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Traditional approaches that use only RGB images have perceptual limitations in complex real-world tasks, including occlusions, lighting changes, sparse textures, and insufficient semantic information.
“Traditional unimodal approaches (such as relying solely on RGB images) often encounter perceptual limitations when dealing with complex tasks in real-world environments-such as occlusions, lighting variations, sparse textures, and insufficient semantic information [1,2,3].”
From Multimodal fusion and vision–language models: A survey for robot vision (2026) · cites this paper as [1]
The Introduction (p. 2) says occlusions make image-based localization and stereo disparity estimation difficult, supporting that example of a perceptual limitation. The paper focuses on stereo-image 3D detection, however, rather than establishing limitations of RGB-only approaches generally; the passages do not support lighting variations, sparse textures, or insufficient semantic information as the stated limitations. (LLM, conf=0.87)
Experimental — accuracy not yet measuredCamera-based single-modal 3D object detection infers 3D information from 2D images; RGB images provide rich visual data but lack direct depth perception, requiring complex estimation, making detection vulnerable to occlusions, poor angles, and adverse weather.
“Camera-based single-modal 3D object detection [1, 72, 2, 3] infers 3D information from 2D images.While RGB images provide rich visual data, the lack of direct depth perception requires complex estimation techniques, making detection vulnerable to occlusions, poor angles, and adverse weather conditions.”
From Multimodal fusion and vision–language models: A survey for robot vision (2026) · cites this paper as [1]
Split into 2 atoms — 1 confirmed, 1 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
Vision meets robotics: The KITTI dataset
Geiger, A., Lenz, P., Stiller, C., & Urtasun, R. (2013) · The International Journal of Robotics Research · doi:10.1177/0278364913491297 · Cited by 9,913 · see citing works
Computer Vision: A Modern Approach
Forsyth, D. A., & Ponce, J. (2002) · Cited by 3,728 · see citing works
Pyramid Stereo Matching Network
Chang, J.-R., & Chen, Y. (2018) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2018.00567 · Cited by 1,882 · see citing works
DenseFusion: 6D Object Pose Estimation by Iterative Dense Fusion
Wang, C., Xu, D., Zhu, Y., Martín-Martín, R., Lu, C., Fei-Fei, L., & Savarese, S. (2019) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2019.00346 · Cited by 1,202 · see citing works
Wang, Y., Chao, W.-L., Garg, D., Hariharan, B., Campbell, M., & Weinberger, K. Q. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00864 · Cited by 1,154 · see citing works
3D Bounding Box Estimation Using Deep Learning and Geometry
Mousavian, A., Anguelov, D., Flynn, J., & Kosecka, J. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.597 · Cited by 1,129 · see citing works
STD: Sparse-to-Dense 3D Object Detector for Point Cloud
Yang, Z., Sun, Y., Liu, S., Shen, X., & Jia, J. (2019) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00204 · Cited by 875 · see citing works
Stereo R-CNN Based 3D Object Detection for Autonomous Driving
Li, P., Chen, X., & Shen, S. (2019) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2019.00783 · Cited by 597 · see citing works
3D Object Proposals Using Stereo Imagery for Accurate Object Class Detection
Chen, X., Kundu, K., Zhu, Y., Ma, H., Fidler, S., & Urtasun, R. (2017) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2017.2706685 · Cited by 424 · see citing works
Multi-level Fusion Based 3D Object Detection from Monocular Images
Xu, B., & Chen, Z. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00249 · Cited by 358 · see citing works
Accurate Monocular 3D Object Detection via Color-Embedded 3D Reconstruction for Autonomous Driving
Ma, X., Wang, Z., Li, H., Zhang, P., Ouyang, W., & Fan, X. N. (2019) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00695 · Cited by 333 · see citing works
ROI-10D: Monocular Lifting of 2D Detection to 6D Pose and Metric Shape
Manhardt, F., Kehl, W., & Gaidon, A. (2019) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2019.00217 · Cited by 305 · see citing works
Monocular 3D Object Detection with Pseudo-LiDAR Point Cloud
Weng, X., & Kitani, K. (2019) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw.2019.00114 · Cited by 290 · see citing works
RoarNet: A Robust 3D Object Detection based on RegiOn Approximation Refinement
Shin, K., Kwon, Y. P., & Tomizuka, M. (2019) · 2019 IEEE Intelligent Vehicles Symposium (IV) · doi:10.1109/ivs.2019.8813895 · Cited by 221 · see citing works
Stereo Vision-Based Semantic 3D Object and Ego-Motion Tracking for Autonomous Driving
Li, P., Qin, T., & Shen, S. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-01216-8_40 · Cited by 156 · see citing works
Triangulation Learning Network: From Monocular to Stereo 3D Object Detection
Qin, Z., Wang, J., & Lu, Y. (2019) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2019.00780 · Cited by 134 · see citing works
Shi, S., Wang, Z., Shi, J., et al. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1907.03670 · Cited by 99 · see citing works
DenseFusion: 6D Object Pose Estimation by Iterative Dense FusionPreprint
Wang, C., Xu, D., Zhu, Y., et al. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1901.04780 · Cited by 91 · see citing works
Wang, Y., Chao, W., Garg, D., et al. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1812.07179 · Cited by 68 · see citing works
3D Object Proposals using Stereo Imagery for Accurate Object Class DetectionPreprint
Chen, X., Kundu, K., Zhu, Y., et al. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1608.07711 · Cited by 49 · see citing works
Stereo R-CNN based 3D Object Detection for Autonomous DrivingPreprint
Li, P., Chen, X., & Shen, S. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1902.09738 · Cited by 35 · see citing works
Wang, R., Yang, N., Stueckler, J., et al. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1904.10097 · Cited by 9 · see citing works
Flagged first, then by citations
Research works printed but not linked: 1
- You, Y.; Wang, Y.; Chao, W.-L.; Garg, D.; Pleiss, G.; Hariharan, B.; Campbell, M.; and Weinberger, K. Q. 2019. Pseudo-lidar++: Accurate depth for 3d object detection in autonomous driving. arXiv preprint arXiv:1906.06310 . · arXiv
Cited by
Sorted by citations
Disparity-Based Multiscale Fusion Network for Transportation Detection
Chen, J., Wang, Q., Peng, W., Xu, H., Li, X., & Xu, W. (2022) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2022.3161977 · Cited by 166 · see citing works
3D Object Detection from Images for Autonomous Driving: A Survey
Ma, X., Ouyang, W., Simonelli, A., & Ricci, E. (2022) · arXiv · doi:10.48550/arxiv.2202.02980 · Cited by 147 · see citing works
Multimodal fusion and vision–language models: A survey for robot vision
Han, X., Chen, S., Fu, Z., Feng, Z., Fan, L., An, D., Wang, C., Guo, L., Meng, W., Zhang, X., et al. (2026) · Information Fusion · doi:10.1016/j.inffus.2025.103652 · Cited by 41 · see citing works
“Camera-based single-modal 3D object detection [1, 77, 2, 3] infers 3D information from 2D images.”
- Authors
- Xu, Z., Zhang, W., Ye, X., Tan, X., Yang, W., Wen, S., Ding, E., Meng, A., & Huang, L.
- Affiliations
- University of Science and Technology of China (CN)
- Baidu (China) (CN)
- Venue
- Proceedings of the AAAI Conference on Artificial Intelligence
- Published
- 2020-04-03
- Volume
- 34
- Issue
- 07
- Pages
- 12557-12564
- ISSN
- 2159-5399 (ISSN-L)
- 2374-3468
- OpenAlex
- W2997558685
- Cited by
- 72
- References
- 22
- Citations by year
- 2025: 6 · 2024: 12 · 2023: 16 · 2022: 17 · 2021: 14 · 2020: 7
- FWCI
- 4.16
- Type
- conference paper
- Language
- en
- Open access
- diamond
- doi.org
- Licence
- www.aaai.org/ (unspecified)
- Locations
- Proceedings of the AAAI Conference on Artificial Intelligence (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- AAAI Publications (The Association for the Advancement of Artificial Intelligence (AAAI)) (submittedVersion)
- arXiv (Cornell University)
- Funders
- National Natural Science Foundation of China (61602400)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Vision and Imaging
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Engineering › Automotive Engineering › Autonomous Vehicle Technology and Safety
- Keywords
- 3D object detection, stereo vision, disparity estimation, occlusion handling, 3D pose estimation, point cloud reconstruction, camera intrinsic parameters, KITTI dataset, autonomous driving
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
