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 301–400 of 597
Wang, Y., Zhang, T., Wu, X., Li, Q., Li, Y., & Wang, Q. (2025) · Plant Phenomics · doi:10.1016/j.plaphe.2025.100048 · Cited by 6
“Speci fi cally, following [34], we use linear-increasing discretization (LID) to convert the depth map into depth discrete bins.”
SelfDN: Adaptive Self-Denoising for multi-view 3D object detection
Qi, Y., Yang, M., Zhang, Y., Xiong, B., Liu, Y., & Zhang, Z. (2025) · Knowledge-Based Systems · doi:10.1016/j.knosys.2025.113816 · Cited by 5
MMFusion: A Generalized Multi-Modal Fusion Detection Framework
Cui, L., Li, X., Meng, M., & Mo, X. (2023) · IEEE International Conference on Development and Learning (ICDL) · doi:10.1109/icdl55364.2023.10364441 · Cited by 7
“The methods [37], [38] rely on a depth estimation network to extract implicit depth information and thus improve the network’s effectiveness.”
MethodEnhanced Cooperative Perception for Autonomous Vehicles Using Imperfect Communication
Sarlak, A., Alzorgan, H., Boroujeni, S. P. H., Razi, A., & Amin, R. (2024) · International Conference on Distributed Computing in Smart Systems and the Internet of Things (DCOSS-IoT) · doi:10.1109/dcoss-iot61029.2024.00108 · Cited by 6
“Authors in [12] evaluate the communication performance of C-V2X or LTE-V Mode 4, focusing on average Packet Delivery Ratio (PDR) and four types of transmission errors, with validating various transmission parameters and traffic densities.”
BackgroundExploiting Dual-Correlation for Multi-frame Time-of-Flight Denoising
Dong, G., Zhang, Y., Sun, X., & Xiong, Z. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-72670-5_27 · Cited by 5
MonoCAPE: Monocular 3D object detection with coordinate-aware position embeddings
Chen, W., Chen, M., Fang, J., Zhao, H., & Wang, G. (2024) · Computers & Electrical Engineering · doi:10.1016/j.compeleceng.2024.109781 · Cited by 5
Lin, B., Zou, Z., & Shi, Z. (2024) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2024.3493963 · Cited by 5
“Alternatively, categorical depth distribution network (CaDNN) [12] incorporates supervision using depth information obtained directly from lidar during depth estimation, leading to a more precise transformation from perspective view to BEV.”
RepVF: A Unified Vector Fields Representation for Multi-task 3D Perception
Li, C., Han, W., Yin, J., Zhao, S., & Shen, J. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-73411-3_16 · Cited by 5
Cross-Supervised LiDAR-Camera Fusion for 3D Object Detection
Zuo, C. J., Gu, C. Y., Guo, Y., & Miao, X. (2024) · IEEE Access · doi:10.1109/access.2024.3518564 · Cited by 5
“CaDDN [11] interpolates LiDAR-projected points to obtain a sparse depth map, using it to supervise the prediction of depth distribution.”
MonoSG: Monocular 3D Object Detection With Stereo Guidance
Fan, Z., Xu, C., Chu, M., Huang, Y., Ma, Y., Wang, J., Xu, Y., & Wu, D. (2025) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2025.3544487 · Cited by 5
“CaDDN [22] predicts a categorical depth distribution for each pixel, projecting rich contextual feature information into the appropriate depth intervals in 3D space.”
Jia, Y., Sun, Y., Mao, R., Nan, Z., Zhou, S., & Niu, Z. (2025) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2025.3621636 · Cited by 5
“CaDDN [26] first encodes images to the 3D spacebydepthpredictionandthencollapsetothe2DBEVspace.”
ETS-3D: An Efficient Two-Stage Framework for Stereo 3D Object Detection
Ji, C., Liu, G., & Zhao, D. (2022) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2022.103634 · Cited by 5
Diaz-Zapata, M., González, D., Erkent, Ö., Laugier, C., & Dibangoye, J. (2023) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra48891.2023.10160757 · Cited by 7
“CaDDN and BEVDet predict a categorical depth distribution to project image features from the camera plane to the BEV [17], [18].”
BackgroundSemi-Supervised Stereo-Based 3D Object Detection via Cross-View Consensus
Wu, W., Wong, H., & Wu, S. (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/cvpr52729.2023.01676 · Cited by 5
“Following [8, 30], we first project the sparse LiDAR points on the image plane to construct the sparse depth maps, followed by interpolating with the efficient depth completion method [16] to generate the corresponding dense depth maps for annotated data.”
MethodMoby: Empowering 2D Models for Efficient Point Cloud Analytics on the Edge
Li, J., Cai, Y. H., Liu, L., Mao, Y. P., Xue, C. J., & Xu, H. L. (2023) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3581783.3612158 · Cited by 5
“The second one is image-based [17, 41, 47, 58, 62] that uses neural networks to recover the depth information from image planes for 3D bounding boxes, which can result in accuracy gaps compared to the point cloud approach.”
BackgroundOccluBEV: Occlusion Aware Spatiotemporal Modeling for Multi-view 3D Object Detection
Wen, Z., Xu, H. C., Liu, C., Guo, T., Hu, J., He, X., Wang, F., Lou, S., & Fan, H. (2023) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3581783.3613798 · Cited by 5
“The first type of works, such as LSS [25] and CaDDN [27], conduct pixel-wise depth estimation and lift the image feature into 3D space accordingly.”
BackgroundMethodHigh-Order Structural Relation Distillation Networks From LiDAR to Monocular Image 3D Detectors
Yan, W., Xu, L., Liu, H., Tang, C., & Zhou, W. (2023) · IEEE Transactions on Intelligent Vehicles · doi:10.1109/tiv.2023.3341981 · Cited by 8
Application of Dynamic Deformable Attention in Bird’s-Eye-View Detection
Gu, W., Ai, R., Liu, J., Fan, L., Cao, D., & Zhang, K. (2022) · IEEE Journal of Radio Frequency Identification · doi:10.1109/jrfid.2022.3210696 · Cited by 5
Predictive World Models from Real-World Partial Observations
Karlsson, R., Carballo, A., Fujii, K., Ohtani, K., & Takeda, K. (2023) · IEEE International Conference on Mobility, Operations, Services and Technologies (MOST) · doi:10.1109/most57249.2023.00024 · Cited by 6
“Later works introduce probabilistic depth projection [60], categorical depth distribution network [61], and multi-task learning [62].”
BackgroundRendBEV: Semantic Novel View Synthesis for Self-Supervised Bird's Eye View Segmentation
Monteagudo, H. P., Taccari, L., Pjetri, A., Sambo, F., & Salti, S. (2025) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv61041.2025.00062 · Cited by 4
AuxDepthNet: Real-Time Monocular 3D Object Detection with Depth-Sensitive Features
Zhang, R., Choi, H., Jung, D., Anh, P. H. N., Jeong, S.-K., & Zhu, Z. (2025) · Applied Sciences · doi:10.3390/app15137538 · Cited by 4
Is Discretization Fusion All You Need for Collaborative Perception?
Yang, K., Bu, T., Li, L., Chunxu, L., Wang, Y., & Li, D. (2025) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra55743.2025.11128776 · Cited by 4
“To assess the performances of ACCO, we leverage two widely used datasets for evaluation: Dair-v2x [6] (real-world) and OPV2V [5] (simulation scenarios).”
Lyu, Y., Tan, X., Yu, Z., & Fan, Z. (2024) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn60899.2024.10650510 · Cited by 4
MixSSC: Forward-Backward Mixture for Vision-Based 3D Semantic Scene Completion
Wang, M., Ding, Y., Liu, Y., Qin, Y., Li, R., & Tang, Z. (2025) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2025.3527235 · Cited by 4
“Advancement of these approaches [36], [37], [38] used actual point clouds to supervise depth distribution for 2D-to-3D transformation, demonstrating that an accurate depth prediction [39], [40], [41] module can significantly improve model performance.”
MonOri: Orientation-Guided PnP for Monocular 3-D Object Detection
Yao, H., Han, P., Chen, J., Wang, Z., Qiu, Y., Wang, X., Wang, Y., Chai, X., Cao, C., & Wei, J. (2025) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2025.3577618 · Cited by 4
“In contrast, ADD’s teacher network takes the depth map as an additional input and demonstrates a significant performance boost across multiple monocular detectors [10], [41], [50].”
InfluentialFeng, Q., Zhao, C., Liu, P., Zhang, Z., Jin, Y., & Tian, W. (2025) · Sensors · doi:10.3390/s25134040 · Cited by 4
MonoAFKD: Align and frequency cross knowledge distillation for monocular 3D object detection
Peng, X., Meng, Y., Zheng, C., & Wang, M. (2025) · Knowledge-Based Systems · doi:10.1016/j.knosys.2025.114043 · Cited by 4
Distribution Aware VoteNet for 3D Object Detection
Liang, J., An, P., & Ma, J. (2022) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v36i2.20049 · Cited by 4
“CaDNN (Reading et al. 2021) adopts pixel-wise depth distribution learning for monocular 3D object detection.”
D-3DLD: Depth-Aware Voxel Space Mapping for Monocular 3D Lane Detection with Uncertainty
Kim, N., Byeon, M., Ji, D., & Oh, D. (2023) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49357.2023.10096483 · Cited by 4
“We adopt channel concatenation rather than channel collapse in 3D object detection [19].”
MethodCheng, X., Qiu, S., Zou, Z., Pu, J., & Xue, X. (2023) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn54540.2023.10191066 · Cited by 4
“CaDDN [16] learns the categorical depth distribution from sampled point clouds in the bird’s-eye-view (BEV) at the training stage, utilizing point clouds to guide depth features.”
BackgroundBlock-Wise Partner Learning for Model Compression
Zhang, X., Xie, W., Li, Y., Lei, J., Jiang, K., Fang, L., & Du, Q. (2023) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2023.3306512 · Cited by 5
BEVMap: Map-Aware BEV Modeling for 3D Perception
Chang, M., Moon, S., Mahjourian, R., & Kim, J. (2024) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv57701.2024.00725 · Cited by 4
A region feature fusion network for point cloud and image to detect 3D object
Shi, Y., Ma, L., Li, J., Wang, X., & Yang, Y. P. (2024) · IET Collaborative Intelligent Manufacturing · doi:10.1049/cim2.12100 · Cited by 4
Yoneda, K., Shiraki, R., Hariya, K., Inoshita, H., Yanase, R., & Suganuma, N. (2024) · IEEE Intelligent Vehicles Symposium · doi:10.1109/iv55156.2024.10588420 · Cited by 4
Guo, M., Zhang, Z., Jing, L., He, Y., Wang, K., & Fan, H. (2024) · International Journal of Computer Vision · doi:10.1007/s11263-024-02176-7 · Cited by 4
“The other burgeoning branch is query-based framework [39,23,30,20,26], which projects each 3D sampling point in BEV space to multi-view 2D images for visual feature extraction.”
3D Object Detection for Autonomous Driving: A Survey
Jin, J., Liu, W., Ning, Z., Zhao, Q., Cheng, S., & da Hu, J. (2024) · Chinese Control and Decision Conference (CCDC) · doi:10.1109/ccdc62350.2024.10587405 · Cited by 4
“Alternatively, CaDDN [25] employs a linear-increasing discretization method, providing balanced depth estimations.”
MethodVIDF-Net: A Voxel-Image Dynamic Fusion method for 3D object detection
Xiang, X., Li, D., Wang, X., Zhou, X., & Qiao, Y. (2024) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2024.104164 · Cited by 4
Kong, L., Zheng, Q., & Wang, W. (2024) · Measurement Science and Technology · doi:10.1088/1361-6501/ad95aa · Cited by 4
“Monocular depth estimation is widely used in the fields of autonomous driving [2, 3], 3D object detection [4, 5], 3D scene reconstruction [6, 7], and augmented reality [8–10].”
MonoDSSMs: Efficient Monocular 3D Object Detection with Depth-Aware State Space Models
Vu, K., Tran, T. T., & Nguyen, D. D. (2024) · Lecture notes in computer science · doi:10.1007/978-981-96-0972-7_6 · Cited by 4
SS-BEV: multi-camera BEV object detection based on multi-scale spatial structure understanding
Shi, P., Pan, Y., & Yang, A. (2025) · Signal Image and Video Processing · doi:10.1007/s11760-024-03762-5 · Cited by 5
MonoAMP: Adaptive Multi-Order Perceptual Aggregation for Monocular 3D Vehicle Detection
Hu, X., Chen, T., Zhang, W., Ji, G., & Jia, H. (2025) · Sensors · doi:10.3390/s25030787 · Cited by 4
DFNet: A Dual LiDAR–Camera Fusion 3-D Object Detection Network Under Feature Degradation Condition
Ye, T., Liu, R., Min, C., Li, Y., & Li, X. (2025) · IEEE Sensors Journal · doi:10.1109/jsen.2025.3551149 · Cited by 5
“Camera-based approaches primarily infer depth information from images using geometric constraints and shape priors and perform detection tasks based on the results of depth estimation [19], [20].”
Joint stereo 3D object detection and implicit surface reconstruction
Li, S., Huang, X., Liu, Z., & Cheng, K.-T. T. (2024) · Scientific Reports · doi:10.1038/s41598-024-64677-2 · Cited by 4
“70 CaDDN [45] CVPR’21 RGB Monocular 24.”
C2ANet: Cross-Scale and Cross-Modality Aggregation Network for Scene Depth Super-Resolution
Ye, X., Guo, Y., Sun, B., Xu, R., Wang, Z., & Li, H. (2023) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2023.3301240 · Cited by 4
Jiang, Q., & Sun, H. (2023) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros55552.2023.10342368 · Cited by 4
NVAutoNet: Fast and Accurate 360° 3D Visual Perception For Self Driving
Pham, T. T., Maghoumi, M., Jiang, W., Jujjavarapu, B. S. S., Sajjadi, M. S. M., Liu, X., Lin, H., Chen, B.-J., Truong, G., Fang, C., et al. (2024) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv57701.2024.00721 · Cited by 4
“BEV representation is advantageous for fusing data from multiple sensors and timestamps and can be directly used by downstream tasks like prediction, planning, and control.”
BackgroundAugmenting 3-D Object Detection Through Data Uncertainty-Driven Auxiliary Framework
Wang, J., Zhao, S., & Liang, S. (2024) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2024.3398077 · Cited by 4
“D object detection algorithm, CaDDN [39], it achieves a performance boost of 0.80% mAP.”
MethodInfluentialPathFusion: Path-Consistent Lidar-Camera Deep Feature Fusion
Wu, L., Wang, D., Li, M., Xiong, Y., Krishnamoorthi, R., Liu, Q., & Chandra, V. (2024) · International Conference on 3D Vision (3DV) · doi:10.1109/3dv62453.2024.00106 · Cited by 4
“Due to the absence of depth information, 2D-to-3D detection heads are developed to augment 2D features with implicitly or explicitly predicted depth, enabling the generation of 3D bounding boxes [17, 18, 23, 29, 30, 37, 46, 48, 50].”
BackgroundZhao, D., Ji, C., & Liu, G. (2023) · Applied Sciences · doi:10.3390/app13031731 · Cited by 4
DRCFusion: dual-stream radar-camera fusion for object detection in adverse weather conditions
Pan, T., Jiang, H., & He, W. (2025) · Journal of Control and Decision · doi:10.1080/23307706.2025.2499526 · Cited by 5
Real-Time Autonomous Driving: A Unified Single-Stage Multi-Task Network Architecture
Raghavendra, S., Kishore, R., Abhilash, S. K., Nookala, V. M., & Yadav, M. V. (2025) · IEEE Access · doi:10.1109/access.2025.3572107 · Cited by 3
Image-based instance segmentation dense point cloud multimodal 3D object detection
Xu, Y., Zhang, R., Shi, P., Zhou, B., Ou, H., & Wang, R. (2025) · Applied Intelligence · doi:10.1007/s10489-025-06661-5 · Cited by 3
Tran, H. T., & Hong, I. (2025) · IEEE Access · doi:10.1109/access.2025.3605872 · Cited by 3
“Advanced approaches such as MonoDIS [14] and CaDDN [15], offer improvements in object boundary estimation but still suffer from sensitivity to varying lighting conditions and occlusions.”
Boosting Monocular 3D Object Detection via 2D Object Detection and Keypoint Estimation
Wen, L.-H., Yu, G., Tian, X.-X., Zhu, W., Liang, G.-Z., Wang, Y., Liu, J.-X., Liu, S., & Tian, Y. (2025) · Conference on Human System Interactions · doi:10.1109/hsi66212.2025.11142398 · Cited by 3
BELT-Fusion: Bayesian Evidential Late Fusion for Trustworthy V2X Perception
Zhao, Z., Zhao, C., Chen, K., Ji, Y., & Du, Y. (2025) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2025.3625597 · Cited by 3
“The success of deep neural networks brings the promise of solving AV’s core requirement to perceive the surrounding environment from point clouds [1], [2], [3], images [4], [5] or multi-modality data [6], [7].”
Ding, Y., Shi, Y., & Guo, J. (2026) · Multimedia Systems · doi:10.1007/s00530-025-02119-3 · Cited by 3
“…Res-101 0.522 0.484 PETR [19] Res-101 0.502 0.462 PETRv2 [20] Res-101 0.519 0.479 BEVFormer [14] Res-101 0.522 0.481 MLDF-BEV MLDFNs 0.525 0.489 [55] Res-101 0.175 0.165 0.050 0.045 DETR3D [18] Res-101 0.220 0.216 0.055 0.051 PETR [19] Res-101 0.235 0.227 0.057 0.051 BEVFormer [14] of the…”
InfluentialMonoPSTR: Monocular 3-D Object Detection With Dynamic Position and Scale-Aware Transformer
Yang, F., He, X., Chen, W., Zhou, P., & Li, Z. (2024) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2024.3415231 · Cited by 3
Better Monocular 3D Detectors with LiDAR from the Past
You, Y., Phoo, C. P., Diaz-Ruiz, C. A., Luo, K. Z., Chao, W.-L., Campbell, M., Hariharan, B., & Weinberger, K. Q. (2024) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra57147.2024.10610444 · Cited by 3
“Current detectors get around this issue by learning a prior over depth [11], [13], [14], [25], [32], but without any geometric or multi-view information, accurate depth estimation can be challenging.”
BackgroundMethodInfluentialXu, Y., Wang, H., Ji, Z., Zhang, Q., Jia, Q., & Li, X. (2024) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2024.3472783 · Cited by 3
“CaDDN [24] learns categorical depth distributions for each pixel and recovers bounding boxes from the bird’s-eye-view (BEV) projection.”
BackgroundDeep Neural Networks for Accurate Depth Estimation with Latent Space Features
Ahn, H., & Yasir, S. M. (2024) · Biomimetics · doi:10.3390/biomimetics9120747 · Cited by 3
“Reading et al. [26] introduced CaDDN, a fully differentiable method for simultaneous object detection and depth estimation, demonstrating the synergy of these two tasks in unified frameworks.”
A Novel 4D Radar and Image Fusion for 3D Object Detection in Autonomous Driving
Zhao, Y., Jin, Y., Deng, R., Tao, Y., Peng, Z., Zhan, H., & Cheng, H. (2024) · IEEE International Conference on Intelligent Transportation Systems (ITSC) · doi:10.1109/itsc58415.2024.10919966 · Cited by 3
“Based on the different intermediate representations, monocular 3D object detection can be divided into three categories: direct methods: direct methods[12], methods based on depth estimation[13][14][15], and methods based on grids[16][17][18].”
Cost-Effective Real-Time Obstacle Detection and Avoidance for AGVs using YOLOv8 and RGB-D Sensors
Medjaldi, A., Slimani, Y., & Karkar, N. (2025) · Engineering Technology & Applied Science Research · doi:10.48084/etasr.10135 · Cited by 3
OccLoff: Learning Optimized Feature Fusion for 3D Occupancy Prediction
Zhang, J., Ding, Y., & Liu, Z. (2025) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv61041.2025.00306 · Cited by 5
“Drawing inspiration from recent advancements in camera-based perception [4, 27, 35, 37], most existing multi-modal approaches [29,31,42,43] perform a 2D-to-3D view transformation on image features, followed by computationally intensive 3D operations—such as 3D convolutions [29, 42, 43] or KNN…”
MonoDFM: Density Field Modeling-Based End-to-End Monocular 3D Object Detection
Liu, G., Huang, X., & Xie, X. (2025) · IEEE Access · doi:10.1109/access.2025.3563248 · Cited by 3
“CaDDN [41] predicts a depth grid distribution for each pixel, mapping image features into 3D space across appropriate depth intervals to generate frustums.”
Region-aware knowledge distillation between monocular camera-based 3D object detectors
Cheon, S.-G., Shin, H.-J., & Bae, S. (2025) · ICT Express · doi:10.1016/j.icte.2025.04.012 · Cited by 3
Monocular 3D Object Detection Based on Pseudo-LiDAR Point Cloud for Autonomous Vehicles
Wang, Y., Xu, S., Zuo, Z., & Li, Z. (2022) · Chinese Control Conference (CCC) · doi:10.23919/ccc55666.2022.9901642 · Cited by 3
“Finally, the third class of method-s realizes 3D object detection in terms of depth maps [6– 9, 16, 20].”
MethodEnd-to-end BEV Perception Via Homography Matrix
Zhang, P., Li, G., Liu, C., & Ma, J. (2023) · IEEE Information Technology, Networking, Electronic and Automation Control Conference (ITNEC) · doi:10.1109/itnec56291.2023.10082203 · Cited by 3
“From 2D to 3D, LSS[5] predicts the depth distribution of each pixel based on two-dimensional characteristics.”
MethodInfluentialMonocular 3D Object Detection From Comprehensive Feature Distillation Pseudo-LiDAR
Sun, C., Xu, C., Fang, W., & Xu, K. (2023) · IEEE Access · doi:10.1109/access.2023.3313432 · Cited by 3
Zhang, Z., Xu, R., & Tian, Q. (2023) · Mathematics · doi:10.3390/math11204277 · Cited by 4
Sound3DVDet: 3D Sound Source Detection using Multiview Microphone Array and RGB Images
He, Y., Shin, S., Cherian, A., Trigoni, N., & Markham, A. (2024) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv57701.2024.00541 · Cited by 3
Depth Estimation From Surface-Ground Correspondence for Monocular 3D Object Detection
Ji, Y., & Xu, J. (2024) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2024.3411159 · Cited by 3
“Following CaDDN [14], we use one third samples of training sequences to conduct the experiments.”
BackgroundMethodZhu, J., Li, X., Xu, Q., Liu, X., Wang, H., & Jiang, T. (2024) · IEEE Sensors Journal · doi:10.1109/jsen.2024.3462751 · Cited by 3
“CaDNN [39] uses a predicted categorical depth distribution for each pixel to project rich contextual feature information to the appropriate depth interval in 3-”
Every Dataset Counts: Scaling up Monocular 3D Object Detection with Joint Datasets Training
Ma, F., Yan, X., Zhao, G., Xu, X., Liu, Y., Ma, J., & Liu, M. (2024) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros58592.2024.10802623 · Cited by 3
BEVFusion With Dual Hard Instance Probing for Multimodal 3D Object Detection
Kim, T., & Kim, J. (2025) · IEEE Access · doi:10.1109/access.2025.3538866 · Cited by 3
“Models like CaDDN [32] and BEVDet [33] employ Lift-Splat-Shoot [34] to transform 2D multi-view image features into the BEV space, facilitating a more intuitive 3D object detection process.”
MonoDGAE: depth-guided attention and bilateral filtering for robust monocular 3D object detection
Bitwire, G. A., Kakuba, S., Cha, D. W., & Han, D. S. (2025) · Artificial Life and Robotics · doi:10.1007/s10015-025-01048-8 · Cited by 3
Wang, Z., Liu, J., Sun, Y., Zhou, X., Sun, B., Kong, D., Xu, J., Yue, X., & Zhang, W. (2025) · PeerJ Computer Science · doi:10.7717/peerj-cs.2656 · Cited by 2
Zhang, T., Yang, X., Liang, Z., Yang, Y., & Wang, J. (2025) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2025.3541372 · Cited by 2
“CaDDN [7] partitions the depth range into segments and predicts the probability for each segment, followed by the expansion of pixel features into frustum features.”
Bird’s-Eye View Perception Based on Two-Stage LiDAR Compensation in Complex Traffic Scenarios
Cong, P., Deng, M., Zhu, Y., Xiao, Y., & Zhang, X. (2025) · IEEE Sensors Journal · doi:10.1109/jsen.2025.3545292 · Cited by 2
“D approach [6], [7], [8], [9], [10], [11], [12] that explicitly estimates the image depth information to complete the construction of BEV viewpoints.”
Guo, Y., Wang, J., Du, C., Xiong, S., & Chen, Y. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10889616 · Cited by 2
Toward Unified 3D Object Detection via Algorithm and Data Unification
Li, Z., Xu, X., Lim, S.-N., & Zhao, H. (2025) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2025.3574363 · Cited by 2
“This approach benefits downstream tasks,asplanningisalsoperformedinthe3Dspace[22].”
Confidence-Guided LiDAR Depth Completion for Robust 3D Object Detection
Hwang, S., & Lee, K. (2025) · IEEE Access · doi:10.1109/access.2025.3609143 · Cited by 2
“Monocular [36], [37], stereo-based [38], and multi-camera based [39] detectors use only images.”
Surface Height and Uncertainty-Based Depth Estimation for Monocular 3D Object Detection
Ji, Y., & Xu, J. (2025) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2025.3611879 · Cited by 2
“Therefore, several works [32], [34], [36] project the LiDAR data onto the image plane and get pixel-wise depth label.”
InfluentialRadar–Camera Fusion in Perspective View and Bird’s Eye View for 3D Object Detection
Xiao, Y., Chen, X., Wang, Y., & Fu, Z. (2025) · Sensors · doi:10.3390/s25196106 · Cited by 5
“CaDDN [18] builds upon the LSS framework and achieves notable performance in 3D object detection.”
Causality-Inspired Debiasing Learning for Open World Object Detection
Zhao, X., Ma, Y., Lv, C., Wang, J., & Liu, X. (2025) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2025.3618534 · Cited by 2
“T HE significant advancements in deep learning-based object detection [1], [2], [3], [4], [5], [6], [7], [8], [9], [10] have enabled unprecedented development across various applications, including visual reasoning [11], [12], [13], [14], [15], [16], [17] and autonomous driving [18], [19], [20].”
Dual Enhancement on 3D Vision-Language Perception for Monocular 3D Visual Grounding
Yu-zhen, L., Liu, M., Bian, Y., Wang, X., Li, Z., Li, G., & Wang, Y. (2025) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3746027.3755497 · Cited by 2
Iqbal, H., Godoy, J., Martin, B., Al-Kaff, A., & García, F. (2025) · Smart Cities · doi:10.3390/smartcities8060197 · Cited by 2
Disentangling Instance and Scene Contexts for 3D Semantic Scene Completion
Liu, E., Yu, E., Chen, S. Y., & Tao, W. (2025) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51701.2025.02506 · Cited by 2
“Frustum-based methods [22, 23, 29, 30, 40, 51] lift image features into 3D frustums by predicting depth probabilities and then use voxel pooling to generate BEV features.”
Roadside Monocular 3D Detection Prompted by 2D Detection
Ma, Y., Hua, W., Li, Y., & Kong, S. (2026) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv61042.2026.00189 · Cited by 2
VPCNet: Voxel-Point Cascade for 3D Object Detection
Zhang, K., Fan, B., Wang, K., & Sun, H. (2022) · China Automation Congress (CAC) · doi:10.1109/cac57257.2022.10054859 · Cited by 2
“Their main task and difficulty is depth estimation, such as CaDDN [1] and DSGN [2].”
BackgroundWang, H., Ji, X., Peng, K. M., Wang, W., & Wang, S. (2023) · Journal of Electronic Imaging · doi:10.1117/1.jei.32.5.053015 · Cited by 2
Bi-Directional Bird’s-Eye View Features Fusion for 3D Multimodal Object Detection and Tracking
Hu, Z.-K., Jhong, S.-Y., Hwang, H.-W., Lin, S.-H., Hua, K., & Chen, Y.-Y. (2023) · International Automatic Control Conference · doi:10.1109/cacs60074.2023.10326208 · Cited by 2
“To address this, recent methods have emerged that transform RGB pixels into 3D space to generate virtual point clouds using depth prediction techniques [5].”
MethodTowards Fair and Comprehensive Comparisons for Image-Based 3D Object Detection
Ma, X., Wang, Y., Zhang, Y., Xia, Z., Meng, Y., Wang, Z., Li, H., & Ouyang, W. (2023) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.00591 · Cited by 2
“CaDDN [63] uses a better transformation strategy [57] and achieves promising results.”
BackgroundMethodZhu, Y., Xu, R., Tao, C., An, H., Wang, H., Sun, Z., & Lu, K. (2024) · Remote Sensing · doi:10.3390/rs16091621 · Cited by 2
Self-supervised 3D vehicle detection based on monocular images
Liu, H., & Sun, Y. (2024) · Signal Processing Image Communication · doi:10.1016/j.image.2024.117149 · Cited by 2
Park, S. K., Chung, J. H., Pae, D.-S., Kang, T., & Lim, M. T. (2024) · Multimedia Tools and Applications · doi:10.1007/s11042-024-19466-4 · Cited by 2
Robust BEV 3D Object Detection for Vehicles with Tire Blow-Out
Yang, D., Fan, X., Dong, W., Huang, C., & Li, J. (2024) · Sensors · doi:10.3390/s24144446 · Cited by 2
“CaDDN [9] trains the voxel features and then collapses them into BEV features that could be used for object detection.”
MethodGAFB-Mapper: Ground aware Forward-Backward View Transformation for Monocular BEV Semantic Mapping
Zhu, J., Yuan, Y., Yin, Z., Zhou, Y., Li, S., Fang, J., & Xue, J. (2024) · IEEE Intelligent Vehicles Symposium · doi:10.1109/iv55156.2024.10588371 · Cited by 2
“BEVDepth [9] and CaDNN [25] introduce point cloud data to provide explicit supervision for depth prediction, demonstrating the significant impact of depth prediction accuracy on 3D detection tasks.”
BackgroundOpen-Vocabulary Predictive World Models from Sensor Observations
Karlsson, R., Asfandiyarov, R., Carballo, A., Fujii, K., Ohtani, K., & Takeda, K. (2024) · Sensors · doi:10.3390/s24144735 · Cited by 2
“Later works propose probabilistic depth projection [82], categorical depth distribution networks [83], and multi-task learning [84] for BEV generation.”
BackgroundHigh‐order multilayer attention fusion network for 3D object detection
Zhang, B., Zhao, Y., Su, C., & Cao, G. (2024) · Engineering Reports · doi:10.1002/eng2.12987 · Cited by 2
“Pseudo-LiDAR methods 26–29 use deep completion networks to map the original RGB images to 3D scenes, generate pseudo point clouds, first extract point cloud features and generate candidate boxes, and then fuse point cloud features and pseudo point cloud features for each candidate box, which can…”
Deformable Feature Fusion Network for Multi-Modal 3D Object Detection
Guo, K., Gan, T., Ding, Z., & Ling, Q. (2024) · International Conference on Robotics, Artificial Intelligence and Intelligent Control (RAIIC) · doi:10.1109/raiic61787.2024.10670940 · Cited by 2
“The detection head stays the same as the baseline[6, 8].”
BackgroundMethodInfluential
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
