The DOI matches its Crossref record; 66 of the 82 references it cites were found in a registry or an index.
Categorical Depth Distribution Network for Monocular 3D Object Detection
Open accessAbstract
Monocular 3D object detection is a key problem for autonomous vehicles, as it provides a solution with simple configuration compared to typical multi-sensor systems. The main challenge in monocular 3D detection lies in accurately predicting object depth, which must be inferred from object and scene cues due to the lack of direct range measurement. Many methods attempt to directly estimate depth to assist in 3D detection, but show limited performance as a result of depth inaccuracy. Our proposed solution, Categorical Depth Distribution Network (CaDDN), uses a predicted categorical depth distribution for each pixel to project rich contextual feature information to the appropriate depth interval in 3D space. We then use the computationally efficient bird’s-eye-view projection and single-stage detector to produce the final output detections. We design CaDDN as a fully differentiable end-to-end approach for joint depth estimation and object detection. We validate our approach on the KITTI 3D object detection benchmark, where we rank 1stamong published monocular methods. We also provide the first monocular 3D detection results on the newly released Waymo Open Dataset. We provide a code release for CaDDN which is made available.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
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?
In the citing paper
“A popular BEV framework that generates BEV features is based on depth information [46, 32, 34], but this paradigm is sensitive to the accuracy of depth values or the depth distributions.”
From BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformers (2022) · cites this paper as [34]
In the citing paper
“BEV-based approaches have not shown significant advantages over other paradigm in 3D object detections [47, 31, 34].”
From BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformers (2022) · cites this paper as [34]
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946
Microsoft COCO: Common Objects in Context
Lin, T.-Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Dollár, P., & Zitnick, C. L. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-10602-1_48 · Cited by 45,460
Fully convolutional networks for semantic segmentation
Long, J., Shelhamer, E., & Darrell, T. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298965 · Cited by 38,834
PyTorch: An Imperative Style, High-Performance Deep Learning Library
Paszke, A., Gross, S., Massa, F., Lerer, A., Bradbury, J., Chanan, G., Killeen, T., Lin, Z., Gimelshein, N., Antiga, L., et al. (2019) · arXiv · doi:10.48550/arxiv.1912.01703 · Cited by 31,473
Focal Loss for Dense Object Detection
Lin, T.-Y., Goyal, P., Girshick, R., He, K., & Dollar, P. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.324 · Cited by 30,581
Chen, L.-C., Papandreou, G., Kokkinos, I., Murphy, K., & Yuille, A. L. (2018) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2017.2699184 · Cited by 22,392
Chen, L.-C., Papandreou, G., Kokkinos, I., Murphy, K., & Yuille, A. L. (2016) · arXiv · doi:10.48550/arxiv.1606.00915 · Cited by 22,392
Are we ready for autonomous driving? The KITTI vision benchmark suite
Geiger, A., Lenz, P., & Urtasun, R. (2012) · 2012 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2012.6248074 · Cited by 14,897
Focal Loss for Dense Object Detection
Lin, T.-Y., Goyal, P., Girshick, R., He, K., & Dollar, P. (2020) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2018.2858826 · Cited by 9,556
Flagged first, then by citations
16 of the 82 references couldn’t be matched to a known paper.
Research works printed but not linked: 4
- mani, 2020, Monolayout: Amodal scene layout from a single image, WACV
- ma, 2020, Rethinking pseudolidar representation, ECC
- shi, 2020, Distance-normalized unified representation for monocular 3d object detection, ECC
- roddick, 2018, Orthographic feature transform for monocular 3D object detection, BMV
Cited by
Sorted by citations
Items 201–300 of 597
MonoAux: Fully Exploiting Auxiliary Information and Uncertainty for Monocular 3D Object Detection
Li, Z., Zheng, W., Yang, L., Ma, L., Zhou, Y., & Peng, Y. (2024) · Cyborg and Bionic Systems · doi:10.34133/cbsystems.0097 · Cited by 20
3D Object Detection for Autonomous Driving: A Comprehensive Review
Nagiub, A. S., Fayez, M., Khaled, H., & Ghoniemy, S. (2024) · International Conference on Computing and Informatics (ICCI) · doi:10.1109/icci61671.2024.10485120 · Cited by 15
“Another approach for incorporating 3D geometry into image-based networks is to learn a latent depth distribution either explicitly or implicitly through an end-to-end process [9, 37, 38].”
BackgroundMethodA review of 3D object detection based on autonomous driving
Wang, H., Chen, X., Yuan, Q., & Liu, P. (2024) · The Visual Computer · doi:10.1007/s00371-024-03480-6 · Cited by 17
On the Road to AIoT: A Framework for Vehicle Road Cooperation
Annunziata, D., Chiaro, D., Qi, P., & Piccialli, F. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3488855 · Cited by 16
“D object detection [29], [30], while LiDAR converts data into voxel or pillar formats [31], [32], [33].”
Context-Aware 3D Object Detection From a Single Image in Autonomous Driving
Zhou, D., Song, X., Fang, J., Dai, Y., Li, H., & Zhang, L. (2022) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2022.3154022 · Cited by 16
OccFeat: Self-supervised Occupancy Feature Prediction for Pretraining BEV Segmentation Networks
Sirko-Galouchenko, S., Boulch, A., Gidaris, S., Bursuc, A., Vobecký, A., P'erez, P., & Marlet, R. (2024) · 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw63382.2024.00452 · Cited by 26
“The predictive performance of LSS can be further improved with other supervision signals, e.g., depth from Lidar points [42, 61], stereo from time [41, 44, 71] or different depth parameterization [80].”
BackgroundOccFusion: Depth Estimation Free Multi-sensor Fusion for 3D Occupancy Prediction
Zhang, J., Ding, Y., & Liu, Z. (2024) · Lecture notes in computer science · doi:10.1007/978-981-96-0972-7_14 · Cited by 19
MonoDFNet: Monocular 3D Object Detection with Depth Fusion and Adaptive Optimization
Gao, Y., Wang, P., Li, X., Sun, M., Di, R., Li, L., & Hong, W. (2025) · Sensors · doi:10.3390/s25030760 · Cited by 14
TiGDistill-BEV: Multi-View BEV 3D Object Detection via Target Inner-Geometry Learning Distillation
Xu, S., Li, F., Huang, P., Song, Z., & Yang, Z.-X. (2025) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2025.3596322 · Cited by 14
“For 3D object detection, previous methods [22], [24], [45] also introduce additional networks for depth estimation to improve the localization accuracy in 3D space.”
PTA-Det: Point Transformer Associating Point Cloud and Image for 3D Object Detection
Wan, R., Zhao, T., & Zhao, W. (2023) · Sensors · doi:10.3390/s23063229 · Cited by 16
“Similar to CaDDN [27], ResNet-101 [38] was utilized as the backbone to process images and the output of its Block1 was used to collect image features FI ∈ RHF×WF×C, where HF,WF were the height and width of the image feature, and C was the number of feature channels.”
BackgroundMethodInfluentialV2VFusion: Multimodal Fusion for Enhanced Vehicle-to-Vehicle Cooperative Perception
Lei Zhang, A., Wang, B., Wang, Z., & Zhao, Y. (2023) · China Automation Congress (CAC) · doi:10.1109/cac59555.2023.10450676 · Cited by 13
“For multi-view camera data, we employ the CaDNN [11] architecture, which comprises five modules: encoder, depth estimation, voxel transformation, and collapse.”
MethodWeakly Supervised Monocular 3D Detection with a Single-View Image
Jiang, X., Jin, S., Lu, L., Zhang, X., & Lu, S. (2024) · 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/cvpr52733.2024.01000 · Cited by 13
“On the other hand, most existing studies [21, 33, 37, 39] adopt a fully supervised setup which have been facing increasing scalability concern as large-scale 3D boxes are often labor-intensive to collect.”
BackgroundZhou, X., Wang, C., Xie, Q., & Qiu, T. (2024) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2024.3486758 · Cited by 13
“The bottom-up methods [21], [30], [31], [32], [33], [34], represented by Lift-Splat-Shoot (LSS) [35], achieves 2D-3D transformation by explicitly predicting the depth for each pixel and generating a series of 3D frustums, and then a voxel pooling operation is used to perform 3D-BEV transformation.”
Semi-supervised Monocular 3D Object Detection by Multi-view Consistency
Lian, Q., Xu, Y., Yao, W., Chen, Y.-C., & Zhang, T. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-20074-8_41 · Cited by 13
“However, due to the intermediate regularization, KM3D [21] only can be adopted to the keypoint-based approaches, while they are less effective compared to other end-to-end detectors [25, 50, 32].”
MethodProbabilistic instance shape reconstruction with sparse LiDAR for monocular 3D object detection
Ji, C., Wu, H., & Liu, G. (2023) · Neurocomputing · doi:10.1016/j.neucom.2023.01.080 · Cited by 13
Learning Geometry-Guided Depth via Projective Modeling for Monocular 3D Object DetectionPreprint
Zhang, Y., Ma, X., Yi, S., Hou, J., Wang, Z., Ouyang, W., & Xu, D. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2107.13931 · Cited by 13
“The prior methods [5, 25, 24, 10, 31] achieve impressive performance on the KITTI test set by introducing extra data or external networks.”
MethodHazarika, A., Vyas, A., Rahmati, M., & Wang, Y. (2023) · IEEE Access · doi:10.1109/access.2023.3288112 · Cited by 14
“Previous state-of-the-art methods utilize either some external depth estimation networks [17] or assume the prior”
MethodMonoSAID: Monocular 3D Object Detection based on Scene-Level Adaptive Instance Depth Estimation
Xia, C., Zhao, W., Han, H., Tao, Z., Ge, B., Gao, X., Li, K., & Zhang, Y. (2023) · Journal of Intelligent & Robotic Systems · doi:10.1007/s10846-023-02027-6 · Cited by 12
“This differs from certain previous methods [11, 41], which fix intervals by pre-defined discretization.”
InfluentialPredict to Detect: Prediction-guided 3D Object Detection using Sequential Images
Kim, S., Kim, Y.-S., Lee, I.-J., & Kum, D. (2023) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.01655 · Cited by 12
“In recent years, camera-based 3D object detection [29, 33, 35, 38] has gained widespread attention due to the cost-effectiveness of a camera sensor and its high-resolution characteristic.”
BackgroundLin, H., Ming, S., Liu, Y., & Qu, X. (2024) · IEEE Transactions on Intelligent Vehicles · doi:10.1109/tiv.2024.3454288 · Cited by 13
DualBEV: Unifying Dual View Transformation with Probabilistic Correspondences
Li, P., Shen, W., Huang, Q., & Cui, D. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-72907-2_17 · Cited by 12
HENet: Hybrid Encoding for End-to-End Multi-task 3D Perception from Multi-view Cameras
Xia, Z., Lin, Z., Wang, X., Wang, Y., Xing, Y., Qi, S., Dong, N., & Yang, M. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-72973-7_22 · Cited by 12
UCDNet: Multi-UAV Collaborative 3-D Object Detection Network by Reliable Feature Mapping
Tian, P., Wang, Z., Cheng, P., Wang, Y., Wang, Z., Zhao, L., Yan, M., Yang, X., & Sun, X. (2024) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2024.3517594 · Cited by 12
“Categorical depth distribution network (CaDDN) [45] and BEVDet [11] propose a fully differentiable end-to-end network that treats feature mapping as a classification task rather than a regression task to predict possibilities at each point along the perspective ray in 3-”
Real-Time Stereo 3D Car Detection With Shape-Aware Non-Uniform Sampling
Gao, A., Cao, J., Pang, Y., & Li, X. (2023) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2022.3220422 · Cited by 12
“CaDDN [43] estimates a categorical depth distribution for each pixel and combines the depth information with the image features for 3D object detection.”
BackgroundDeep Learning-Based Monocular 3D Object Detection with Refinement of Depth Information
Hu, H., Zhu, M. Q., Li, M., & Chan, K. L. (2022) · Sensors · doi:10.3390/s22072576 · Cited by 13
A Multimodal 3-D Detector With Attention From the Corresponding Modal
Tao, B., Yan, F., Yin, Z., Nie, L., Miao, M., Jiao, Y., & Lei, C. (2023) · IEEE Sensors Journal · doi:10.1109/jsen.2023.3252178 · Cited by 11
“Inspired by categorical depth distribution network (CaDDN) [26], we develop an image voxel generator module, whose working flow is presented in Fig.”
BackgroundMethod3D Video Object Detection with Learnable Object-Centric Global Optimization
He, J., Chen, Y., Wang, N., & Zhang, Z. (2023) · 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/cvpr52729.2023.00494 · Cited by 11
GRAMO: geometric resampling augmentation for monocular 3D object detection
Guan, H., Song, C., & Zhang, Z. (2024) · Frontiers of Computer Science · doi:10.1007/s11704-023-3242-2 · Cited by 11
Multimodal 3D Object Detection Based on Sparse Interaction in Internet of Vehicles
Li, H., Ge, T., Bai, K., Nie, G., Xu, L., Ai, X., & Cao, S. (2024) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2024.3357492 · Cited by 13
“[14]similarlypredictthedepthdistributionfor eachpixel,projectingsemanticinformationinto3Dspaceand converting it into BEV features to achieve the detection task; Guo et al. [15] propose using features from the LiDAR-based detector as guidance to improve the detection accuracy.”
A Robust Diffusion Modeling Framework for Radar Camera 3D Object Detection
Wu, Z., Wu, Y., Wang, X., Gan, Y., & Pu, J. (2024) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv57701.2024.00325 · Cited by 12
“Based on a single image input, recent approaches adopted geometric constraint regularization [3, 24, 30, 34, 40, 63] and depth estimation interaction [37, 38, 47] to assist 3D object detection.”
MethodAlleviating Foreground Sparsity for Semi-Supervised Monocular 3D Object Detection
Zhang, W., Liu, D., Ma, C., & Cai, W. (2024) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv57701.2024.00737 · Cited by 11
Improving Distant 3D Object Detection Using 2D Box Supervision
Yang, Z., Yu, Z., Choy, C., Wang, R., Anandkumar, A., & Álvarez, J. M. (2024) · 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/cvpr52733.2024.01407 · Cited by 11
Bi-SSC: Geometric-Semantic Bidirectional Fusion for Camera-Based 3D Semantic Scene Completion
Xue, Y., Li, R., Wu, F., Tang, Z., Li, K., & Duan, M. (2024) · 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/cvpr52733.2024.01902 · Cited by 11
“To transform image features into BEV features, researchers have employed the LSS [33] framework along with subsequent investigations [14, 35, 56, 60].”
MethodWan, Y., Lv, P., Sun, L., Yang, Y., & Hao, J. (2024) · Knowledge-Based Systems · doi:10.1016/j.knosys.2024.112577 · Cited by 11
IFTR: An Instance-Level Fusion Transformer for Visual Collaborative Perception
Wang, S., Bin, L., Xiao, X., Xiang, Z., Shan, H., & Liu, E. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-73021-4_8 · Cited by 11
Large-kernel spatially parallel feature fusion for monocular 3D perception in autonomous driving
Jiao, R., Zhang, J., Li, C., & Hu, L. (2026) · Knowledge-Based Systems · doi:10.1016/j.knosys.2026.115998 · Cited by 10
YOLOv7-3D: A Monocular 3D Traffic Object Detection Method from a Roadside Perspective
Ye, Z., Zhang, H., Gu, J., & Li, X. (2023) · Applied Sciences · doi:10.3390/app132011402 · Cited by 10
From a Bird's Eye View to See: Joint Camera and Subject Registration without the Camera Calibration
Qian, Z., Han, R., Feng, W., & Wang, S. (2024) · 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/cvpr52733.2024.00088 · Cited by 10
“In BEV, we can obtain the global spatial layout and trajectories of all the persons in the scene without mutual occlusion, which has some typical scenarios including the automatic driving [11, 28, 29, 31, 41], outdoor human detection [4, 10] and complementary-view crowd analysis [14, 18].”
MethodNavigation Instruction Generation with BEV Perception and Large Language Models
Fan, S., Liu, R., Wang, W., & Yang, Y. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-72670-5_21 · Cited by 17
MonoTTA: Fully Test-Time Adaptation for Monocular 3D Object Detection
Lin, H., Zhang, Y., Niu, S., Cui, S. R., & Li, Z. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-72784-9_6 · Cited by 10
Yang, T., Qian, Y., Yan, W., Wang, C., & Yang, M. Z. (2024) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2024.3492289 · Cited by 11
“The forward projection method [4], [27], [34], [35], exemplified by LSS [4], predicts the pixel-wise depth distribution by a dedicated depth estimation network, eventually obtaining BEV features through grid sampling.”
Exploiting Ground Depth Estimation for Mobile Monocular 3D Object Detection
Zhou, Y., Liu, Q., Zhu, H., Li, Y., Chang, S., & Guo, M. (2025) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2025.3529084 · Cited by 10
“Such prior knowledge can be derived in various ways, such as a depth map generated by LiDAR point cloud (or Pseudo-LiDAR) [8], [34], [47], [59], monocular depth predictors [17], [33], [45], or disparity map generated by stereo cameras [63].”
AVCPNet: An AAV-Vehicle Collaborative Perception Network for 3-D Object Detection
Wang, Y., Wang, Z., Cheng, P., Tian, P., Yuan, Z., Tian, J., Wang, W., & Zhao, L. (2025) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2025.3546669 · Cited by 10
“D visual detection [34], [35], [36], [37] and play a crucial role in the modeling process, where accurate depth values help mitigate distortions.”
Adaptive learning point cloud and image diversity feature fusion network for 3D object detection
Yan, W., Liu, S., Liu, H., Yue, G., Wang, X., Song, Y., & Xu, J. (2023) · Complex & Intelligent Systems · doi:10.1007/s40747-023-01295-x · Cited by 11
“CaDDN [27] utilized classified depth distribution and projection geometry to generate high-quality Bird’s Eye View (BEV) scene representations from a single image.”
InfluentialSeaBird: Segmentation in Bird's View with Dice Loss Improves Monocular 3D Detection of Large Objects
Kumar, A., Guo, Y., Huang, X., Ren, L., & Liu, X. (2024) · 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/cvpr52733.2024.00978 · Cited by 10
“Advancements include introducing new architectures [33, 88, 105], equivariance [11, 43], losses [4, 14], uncertainty [41, 63] and incorporating auxiliary tasks such as depth [71, 115], NMS [42, 56, 87], corrected extrinsics [118], CAD models [10, 45, 60] or LiDAR [81] in training.”
MethodGao, H., Yu, X., Xu, Y., Ran, Q., & Hussain, W. (2024) · ACM Transactions on Autonomous and Adaptive Systems · doi:10.1145/3703458 · Cited by 9
“Table 3 reveals that our method outperformed the other methods on the cyclist category, outperforming the runner-up CaDDN approach[18] by 3.23%, 2.19%, and 1.37% under the easy, moderate and hard settings, respectively.”
InfluentialMoVis: When 3D Object Detection Is Like Human Monocular Vision
Wang, Z., Yi, J., Chen, A., & Han, G. (2025) · IEEE Transactions on Image Processing · doi:10.1109/tip.2025.3544880 · Cited by 9
“CaDDN [34] discretizes continuous depth into intervals and projects contextual features into appropriate 3D depth bins by predicting the depth distribution of each pixel.”
CARLA Drone: Monocular 3D Object Detection from a Different Perspective
Meier, J., Scalerandi, L., Dhaouadi, O., Kaiser, J., Araslanov, N., & Cremers, D. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-85187-2_9 · Cited by 9
Yu, H., Zhang, X., Zhou, X., & Chan, S. (2025) · Neural Networks · doi:10.1016/j.neunet.2025.107818 · Cited by 10
HV-BEV: Decoupling Horizontal and Vertical Feature Sampling for Multi-View 3D Object Detection
Wu, D., Yang, F., Xu, B., Liao, P., Zhao, W., & Zhang, D. (2025) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2025.3590464 · Cited by 9
“LSS [24] is a foundational work in this category, predicting the discrete depth distribution [25] of pixel points and, through a novel outer product with 2D image features, lifting these features into a frustum-like 3D perspective space to simulate feature point distribution along the line of sight.”
Wang, S., & Zheng, J. Y. (2023) · Frontiers in artificial intelligence and applications · doi:10.3233/faia230555 · Cited by 9
“According to the input resources, the existing 3D object detectors can be divided into four categories: LiDAR point cloud-based [19], stereo image-based [13], monocular imagebased [31, 42] and multi-modality-based methods [36].”
BackgroundGPro3D: Deriving 3D BBox from ground plane in monocular 3D object detection
Yang, F., Xu, X., Chen, H., Guo, Y., He, Y., Ni, K., & Ding, G. (2023) · Neurocomputing · doi:10.1016/j.neucom.2023.126894 · Cited by 10
Translating Images to Road Network: A Non-Autoregressive Sequence-to-Sequence Approach
Lu, J., Li, H., Peng, R., Wen, F., Cai, X., Zhang, W., Xu, H., & Zhang, L. (2023) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.00009 · Cited by 13
“To learn ego-car feature learning from onboard cameras, [56, 43] project image features to ego-car coordinate based on depth estimation.”
BackgroundSze, S., & Kunze, L. (2024) · IEEE Intelligent Vehicles Symposium · doi:10.1109/iv55156.2024.10588696 · Cited by 9
Bird’s Eye View Perception for Autonomous Driving
Du, J., Su, S., Fan, R., & Chen, Q. (2023) · Advances in computer vision and pattern recognition · doi:10.1007/978-981-99-4287-9_10 · Cited by 9
ODD-M3D: Object-Wise Dense Depth Estimation for Monocular 3-D Object Detection
Park, C., Kim, H., Jang, J., & Paik, J. (2024) · IEEE Transactions on Consumer Electronics · doi:10.1109/tce.2024.3366763 · Cited by 8
“CaDDN predicts the depth distribution and the feature extraction in parallel and uses the estimated depth distribution as a frustum feature grid [17].”
MethodLiDAR-Camera Cross Fusion Network Toward 3-D Object Detection in Self-Driving
Chen, B., Shen, H., Zhao, Z., Yu, L., & Zhao, Y. (2024) · IEEE Sensors Journal · doi:10.1109/jsen.2024.3425390 · Cited by 8
“D object detection meth-ods can be divided into three categories: LiDAR-based [2], [3], [4], [5], [6], [7], [8], [9], camera-based [10], [11], [12], [13], [14], [15], [16], [17], and fusion-based methods [18], [19], Manuscript received 14 March 2024; accepted 22 June 2024.”
MCCA-MOT: Multimodal Collaboration-Guided Cascade Association Network for 3D Multi-Object Tracking
Li, H., Liu, H., Du, Z., Chen, Z., & Tao, Y. (2024) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2024.3489234 · Cited by 8
“4, we convert image features into frustum features G based on CaDDN [32] and HMFI [33].”
CT3D++: Improving 3D Object Detection with Keypoint-Induced Channel-wise Transformer
Hua, X.-S., Cai, S., Zhao, N., Deng, B., Liang, Q., Zhao, M.-J., & Ye, J. (2025) · International Journal of Computer Vision · doi:10.1007/s11263-025-02404-8 · Cited by 10
“Alternatively, several other studies, including D 4 LCN Ding et al. (2020), Liu et al. (2021b), CaDDN Reading et al. (2021), and MonoDETR Zhang et al. (2023), opt to combine depth maps with image features to produce the final 3D detection results.”
Advancements in 3-D object detection: A comprehensive review
Wu, W.-J., Appiah, I., & Hu, R. (2025) · Journal of King Saud University - Computer and Information Sciences · doi:10.1007/s44443-025-00213-0 · Cited by 9
Your Camera Improves Your Point Cloud Compression
Lin, Y., Xu, T., Zhu, Z., Li, Y., Wang, Z., & Wang, Y. (2023) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49357.2023.10095556 · Cited by 8
“Inspired by previous success on monocular 3D object detection and monocular depth estimation [12], we estimate depth information by predicting discrete probability distributions at different pixels through visible images.”
BackgroundBEVUDA: Multi-geometric Space Alignments for Domain Adaptive BEV 3D Object Detection
Liu, J., Zhang, R., Li, X., Chi, X., Chen, Z., Lu, M. F., Guo, Y., & Zhang, S. (2024) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra57147.2024.10610096 · Cited by 8
“According to previous work [2], [35], [19], [20], ResNet-50 and ResNet-101 [60] serve as backbone to extract image features respectively.”
MethodActFormer: Scalable Collaborative Perception via Active Queries
Huang, S., Zhang, J., Li, Y., & Feng, C. (2024) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra57147.2024.10610997 · Cited by 8
“[29, 30, 31] estimate the depth distribution then project the 2D image features to 3D space to get the BEV voxel features.”
BackgroundOA-DET3D: Embedding Object Awareness As A General Plug-in for Multi-Camera 3D Object Detection
Chu, X., Deng, J., Ji, J., Zhang, Y., Li, H., & Zhang, Y. (2025) · International Journal of Computer Vision · doi:10.1007/s11263-025-02544-x · Cited by 8
DST3D: DLA-Swin Transformer for Single-Stage Monocular 3D Object Detection
Wu, Z., Jiang, X., Xu, R., Lu, K., Zhu, Y., & Wu, M. (2022) · IEEE Intelligent Vehicles Symposium · doi:10.1109/iv51971.2022.9827462 · Cited by 8
Depth Is All You Need for Monocular 3D Detection
Park, D., Li, J., Chen, D., Guizilini, V., & Gaidon, A. (2023) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra48891.2023.10160483 · Cited by 8
Chen, Y., Dong, S., Hu, B., Liu, Q., & Qu, Y. (2023) · Measurement Science and Technology · doi:10.1088/1361-6501/ad0f08 · Cited by 9
“Classical semantic segmentation algorithms based on convolution neural network include FCN [22], PSPNet [23], UNet [24, 25], DeepLab series [5, 26–28], Swin-Transformer-Segment [29, 30] and so on.”
Rlm-tracking: online multi-pedestrian tracking supported by relative location mapping
Ren, K., Hu, C., & Xi, H. (2024) · International Journal of Machine Learning and Cybernetics · doi:10.1007/s13042-023-02070-7 · Cited by 7
Temporal Feature Fusion for 3D Detection in Monocular Video
Cheng, H., Peng, L., Yang, Z., Lin, B., He, X., & Wu, B. (2024) · IEEE Transactions on Image Processing · doi:10.1109/tip.2024.3378475 · Cited by 7
“Some works use depth to project image features to 3D space [47], or convert RGB image as a pseudo-LiDAR point cloud [36], [60].”
BackgroundMethodInfluentialDMFusion: LiDAR-camera fusion framework with depth merging and temporal aggregation
Yu, X., Lu, K., Yang, Y., & Ou, L. (2024) · Applied Intelligence · doi:10.1007/s10489-024-05627-3 · Cited by 7
“The third set of methods, inspired by LiDAR-based detectors such as LSS [7] and CaDDN [27], explicitly estimated the depth distribution of a 2D image.”
BEVUDA++: Geometric-Aware Unsupervised Domain Adaptation for Multi-View 3D Object Detection
Zhang, R., Liu, J., Li, X., Chi, X., Wang, D., Du, L., Du, Y., & Zhang, S. (2024) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2024.3523049 · Cited by 7
“…3D Object Detection has become a cornerstone in autonomous driving and machine scene comprehension, with two leading approaches: Single-view [3], [8], [11], [18], [37], [43], [46], [57], [66], [77] and Multi-view [14], [26], [27], [30], [32], [33], [34], [38], [41], [42], [48], [51], [68].”
View-to-label: Multi-view consistency for self-supervised monocular 3D object detection
Mouawad, I., Brasch, N., Manhardt, F., Tombari, F., & Odone, F. (2025) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2025.104320 · Cited by 7
SGV3D: Toward Scenario Generalization for Vision-Based Roadside 3D Object Detection
Yang, L., Zhang, X., Li, J., Wang, L., Zhang, C., Ju, L., Li, Z., Shen, Y., Lv, C., & Wang, H. (2025) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2025.3569399 · Cited by 12
“Voxel Pooling [47] converts the 3D wedge-shaped features into masked BEV features F be v mask along the height direction.”
Multi-Modal BEV Enhancement Fusion for 3D Object Detection in Autonomous Driving
Chen, M., Yang, M., Zhang, Y., Han, T., Li, X., Zhao, H., & Liu, P. (2025) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2025.3589581 · Cited by 7
“This limitation has spurred interest in Bird’s-Eye View (BEV) methodologies [11], [12], [13], [14], [15], [16], [17], [18], [19], which offer a unified coordinate system for fusing data from various camera angles, effectively alleviating redundancy issues.”
Song, Q., Hu, Q., Zhang, C., Chen, Y., & Huang, R. (2024) · IEEE Transactions on Image Processing · doi:10.1109/tip.2024.3352808 · Cited by 10
“There are two mainstream approaches to multi-camera 3D perception tasks: 1) Depth-based methods, which estimate pseudo-depth to project multi-view image features into 3D space [1], [14], [15], [16].”
MonoWAD: Weather-Adaptive Diffusion Model for Robust Monocular 3D Object Detection
Oh, Y., Kim, H.-I., Kim, S. T., & Kim, J. U. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-72684-2_19 · Cited by 7
Adv3D: Generating 3D Adversarial Examples for 3D Object Detection in Driving Scenarios with NeRF
Li, L., Lian, Q., & Chen, Y.-C. (2024) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros58592.2024.10801323 · Cited by 7
Monomm: a multi-scale mamba-enhanced network for real-time monocular 3D object detection
Fu, Y., Xu, Z., Fu, J., Xue, H., Tan, S., Li, L., & Qing, S. (2025) · The Journal of Supercomputing · doi:10.1007/s11227-025-06946-z · Cited by 7
“Regarding 3D detection for cyclists, we also achieved results comparable to CaDDN [43] in the easy difficulty and surpassed other methods in the moderate and hard difficulties.”
InfluentialA Survey for 3D Object Detection Algorithms from Images
Lee, H. L., Kim, Y., & Kim, B. (2022) · Journal of Multimedia Information System · doi:10.33851/jmis.2022.9.3.183 · Cited by 7
Consistent GT-Proposal Assignment for Challenging Pedestrian Detection
Luo, Y., Zhao, M., Sun, J., Zhai, G., & Zhang, C. (2024) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2024.3401549 · Cited by 6
Sparse Points to Dense Clouds: Enhancing 3D Detection with Limited LiDAR Data
Kumar, A., Chen, C., Mian, A. S., Lobo, N., & Shah, M. A. (2024) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros58592.2024.10801818 · Cited by 7
“Another approach constructs 3D feature volumes from 2D features to improve the 3D detection capability [16], [25].”
Wang, W., Han, J., Jiang, Z., Zhou, Z., & Wu, Y. (2025) · IEEE Internet of Things Journal · doi:10.1109/jiot.2025.3525899 · Cited by 8
“CaDDN [10] predicted each pixel’s depth distribution, using LiDAR point cloud data as depth labels for depth estimation supervision.”
Oladele, D. A., Markus, E. D., & Abu-Mahfouz, A. M. I. (2025) · AI · doi:10.3390/ai6040082 · Cited by 7
Deep learning-based object detection for environmental monitoring using big data
Lin, W., Li, T., & Li, X. (2025) · Frontiers in Environmental Science · doi:10.3389/fenvs.2025.1566224 · Cited by 11
Ev-3DOD: Pushing the Temporal Boundaries of 3D Object Detection with Event Cameras
Cho, H., Kang, J., Kim, Y.-H., & Yoon, K. (2025) · 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/cvpr52734.2025.02533 · Cited by 11
“To overcome this challenge, several studies [7, 70, 76, 83, 98, 115] have employed depth estimation techniques to generate pseudo 3D point clouds or elevate 2D features into 3D space for object detection, alongside proposals for transformer-based architectures [35, 99, 121] that utilize 3D object…”
Zfusion: an Effective Fuser of Camera and 4D Radar for 3D Object Perception in Autonomous Driving
Yang, S., Zhan, T., Qiao, S., Gong, J., Yang, Q. X., Wang, J., & Lu, Y. (2025) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw67362.2025.00362 · Cited by 12
“Another stream of research [27, 30, 33, 46] obtains BEV features via depth prediction networks and feature converters.”
You Only Look Bottom-Up for Monocular 3D Object Detection
Xiong, K., Zhang, D., Liang, D., Liu, Z., Yang, H., Dikubab, W., Cheng, J., & Bai, X. (2023) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2023.3313053 · Cited by 6
“Depth estimation-based meth-ods [34], [33], [24], [51], [45] mainly estimate pixel-level depth or depth distributions to transform the image features into 3D features.”
BackgroundMethodMonocular-GPS Fusion 3D object detection for UAVs
Ren, S., Zhao, W., Zhang, A., Zhang, B., & Han, B. (2024) · Knowledge-Based Systems · doi:10.1016/j.knosys.2024.112134 · Cited by 7
DSAV: A Deep Sparse Acceleration Framework for Voxel-Based 3-D Object Detection
Fang, H., Tan, Y., Ren, A., Zhuang, W., Hua, Y., Qin, Z., & Liu, D. (2024) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2024.3437334 · Cited by 6
Infrastructure-Assisted Collaborative Perception in Automated Valet Parking: A Safety Perspective
Jia, Y., Zhang, J., Lu, S., Fan, B., Mao, R., Zhou, S., & Niu, Z. (2024) · IEEE Vehicular Technology Conference · doi:10.1109/vtc2024-spring62846.2024.10683664 · Cited by 6
“For camera images, we introduce the Categorical Depth Distribution Network (CaDDN) [17] to achieve camera-to-BEV transformation.”
MethodCollabGAT: Collaborative Perception Using Graph Attention Network
Ahmed, A. N., Mercelis, S., & Anwar, A. (2024) · IEEE Access · doi:10.1109/access.2024.3468432 · Cited by 6
“This work utilizes a constant geometric transformation to transform the characteristics of the 2D pseudo-image to a BEV, aligning with the reliable CaDDN [36] approach.”
MethodSparseDet: Towards efficient multi-view 3D object detection via sparse scene representation
Li, J., Yang, L., Shi, Z., Chen, Y., Jin, Y., Akiyama, K., & Xu, A. (2024) · Advanced Engineering Informatics · doi:10.1016/j.aei.2024.102955 · Cited by 6
Multitask Collaborative Perception for Vehicle-to-Everything Considering Impaired Communication
Yan, F., Tao, B., Zheng, N., Nie, L., Li, Q., & Yin, Z. (2025) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2025.3548801 · Cited by 6
“With the rapid development of AI methods and advanced sensor technologies [1], [2], multiple perceptional tasks such as object detection [3], [4], object tracking [5], [6], and map segmentation, [7], [8] have shown impressive progress.”
Diffusion-Based Particle-DETR for BEV Perception
Nachkov, A., Paudel, D. P., Danelljan, M., & Gool, L. V. (2025) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv61041.2025.00270 · Cited by 9
“Implicit projection, where depth is not estimated explicitly, can be achieved by utilizing self-attention to look up the past BEV and cross-attention to look up the current image features [28, 29, 34, 49, 67, 71].”
MonoDINO-DETR: Depth-Enhanced Monocular 3D Object Detection Using a Vision Foundation Model
Kim, J., Moon, S., Nahl, S., & Shim, D. H. (2025) · IEEE Intelligent Vehicles Symposium · doi:10.1109/iv64158.2025.11097713 · Cited by 6
“Both models utilize depth distributions for each pixel, following the approach introduced in CaDDN [12] for supervision.”
One Training for Multiple Deployments: Polar-based Adaptive BEV Perception for Autonomous Driving
Yang, H., Bai, X., Zhu, X., & Ma, Y. (2023) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra48891.2023.10161552 · Cited by 6
“In addition, depth-based methods [31, 9, 42, 28, 7, 2] estimate depth distribution explicitly or implicitly and use it as the bridge to generate BEV feature from multi-view image”
MethodCI3D: Context Interaction for Dynamic Objects and Static Map Elements in 3D Driving Scenes
Cai, F., Chen, H., & Deng, L. (2023) · IEEE Transactions on Image Processing · doi:10.1109/tip.2023.3340607 · Cited by 7
“Thanks to KITTI [11] benchmark, great advancement has been achieved in monocular 3D object detection research [12], [13], [14], [15], [16], [17] for the last few years. nuScenes [18], another large-scale dataset released subsequently, becomes popular with the computer vision community due to the…”
BackgroundScene Representation in Bird’s-Eye View from Surrounding Cameras with Transformers
Yun, Z., Zhang, Y., Gong, Z., & Zhu, H. (2022) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw56347.2022.00497 · Cited by 6
“CaDDN [21] follows the project mechanism in [20].”
BackgroundLi, S., Liu, Z., Shen, Z., & Cheng, K.-T. T. (2022) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v36i2.20026 · Cited by 6
Bi-Att3DDet: Attention-Based Bi-Directional Fusion for Multi-Modal 3D Object Detection
Gao, X., Zhao, Y., Wang, Y., Shang, J., Zhang, C., & Wu, G. (2025) · Sensors · doi:10.3390/s25030658 · Cited by 7
Metadata
- Authors
- Reading, C., Harakeh, A., Chae, J., & Waslander, S. L.
- Affiliations
- University of Toronto (CA)
- Venue
- 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
- Publisher
- IEEE
- Published
- 2021-06
- Pages
- 8551-8560
- ISSN
- 1063-6919 (ISSN-L)
- OpenAlex
- W3133598367
- Cited by
- 595
- References
- 82
- Citations by year
- 2026: 57 · 2025: 143 · 2024: 173 · 2023: 136 · 2022: 73 · 2021: 13
- FWCI
- 32.30
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- Locations
- IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings (publishedVersion)
- arXiv (Cornell University) (submittedVersion, cc-by)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University) (cc-by)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Engineering › Automotive Engineering › Autonomous Vehicle Technology and Safety
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Vision and Imaging
- Keywords
- monocular 3D object detection, depth estimation, single-stage detectors, autonomous driving, Waymo Open Dataset, KITTI dataset
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
