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
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 501–597 of 597
Uncertainty-aware Monocular 3D Object Detection for Safe Driving Assistance
Ida, M., & Matsushima, K. (2024) · International Symposium on Intelligent Signal Processing and Communication Systems (ISPACS) · doi:10.1109/ispacs62486.2024.10869219
MoPoSD: monocular 3D object detector with potential for specialized exploration in depth
Lin, J.-Y., Liu, S., & Wang, Y. (2025) · International Conference on Graphics and Image Processing (ICGIP) · doi:10.1117/12.3057826
“According to [37], we use LID(Linear Increasing Discretization)[11] to discretize the depth ground truth into 𝐿 + 1 depth bins for depth representation, where the ordered 𝐿 bins represent the foreground depth and the last one represents the background, and limit the foreground depth value to [ 𝑑…”
Multi-KeyPoint Depth Fusion for Monocular 3D Object Detection
Zhang, S., Sun, M., Cui, X., & Huang, Y. W. (2024) · International Conference on Communication Networks and Machine Learning (CNML) · doi:10.1145/3701047.3701069
“Compared to methods such as CaDDN [26] and D4LCN [30], which use additional data, our method shows slight improvements in the moderate and hard levels, though it is somewhat inferior at the easy level.”
LSTV-V2V: A Large-Scale Traffic Virtual Dataset for Vehicle-to-Vehicle Cooperative Perception
Suo, X., Zhang, H., Xu, F., & Li, Y. (2024) · IEEE International Conference on Intelligent Transportation Systems (ITSC) · doi:10.1109/itsc58415.2024.10920245
“(1) Image-based 3D object detection refers to the method that uses monocular/stereo images as input to predict the 3D objects [7]–[9].”
Monocular 3D Object Detection for Robotics
Nurdin, F. R., Sadida, Q., Haq, Q. M. U., Islam, N. U., & Haa, M. A. (2024) · Beyond Technology Summit on Informatics International Conference (BTS-I2C) · doi:10.1109/bts-i2c63534.2024.10942080
“Point clouds are three-dimensional representations of the environment generated by Lidar sensors by mapping the distance of each point in the environment relative to the sensor [15].”
RO‐BEV: Towards Robust BEV Feature Enhancement for Surround‐View Instance Prediction
Yang, Z., Lü, T., Wang, J., Lang, X., Fu, S., & Han, J. (2025) · Electronics Letters · doi:10.1049/ell2.70207
HP-BEV: a bird’s-eye view segmentation model based on spatial height modeling
Sang, H., & Cheng, H. (2025) · Journal of Electronic Imaging · doi:10.1117/1.jei.34.3.033019
MonoDDT: Monocular 3D Object Detection Enhanced by Data-Driven Techniques
Fan, Z., Xu, C., Huang, Y., Ma, Y., Wang, J., Xu, Y., Qian, M., & Wu, D. (2025) · International Conference on Software Engineering and Computer Science (CSECS) · doi:10.1109/csecs64665.2025.11009909
“CaDDN [26] utilizes point cloud data to generate pseudo-depth maps for supervised learning, predicting the classification depth distribution for each pixel to assist in 3D object detection.”
Accelerating point cloud analytics on resource-constrained edge devices
Li, J., Cai, Y. H., Liu, L., Mao, Y. P., Xue, C. J., & Xu, H. (2025) · Computer Networks · doi:10.1016/j.comnet.2025.111382
GaussBEV: Multi-head Gaussian Feature Grouping for Multi-view 3D Object Detection
Zhang, J., Zhou, S., Huang, Y., Dong, J., Fu, J., & Zheng, N. (2025) · IFIP advances in information and communication technology · doi:10.1007/978-3-031-96239-4_19
FC-MonoDETR: A Monocular 3D Object Detection Network Based on Foreground Constraint
Xiao, D., Yu, D., Wang, Y., Xiao, J., Wang, Y., & Liu, L. (2025) · International Conference on Multimedia Retrieval · doi:10.1145/3731715.3733335
“Advanced techniques include CaDDN’s depth distribution prediction [25] and MonoPGC’s geometric context optimization [34].”
Improved Virtual Convolution for 3D Object Detection Using Multimodal Data
Kunjumon, R., Mohamed, S. M. K., & Alqumsan, A. A. (2025) · Communications in computer and information science · doi:10.1007/978-981-96-6957-8_3
Modal Mimicking Knowledge Distillation for monocular three-dimensional object detection
Yang, M., Qi, Y., Xiong, B., Zhang, Z., & Liu, Y. (2025) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2025.111821
Deep Learning Based 3D Target Detection
Ge, Z., Li, F., & Li, P. L. (2025) · International Conference on Autonomous Driving and Intelligent Sensing Technology (ADIST) · doi:10.1145/3744439.3744444
“The KITTI dataset collection are all real autopilot data images, each set contains view and corresponding point cloud data, and the division of training data by Chen and others is used, in which there are 3712 sets of training set and 3769 sets of validation set [8].”
OCC-Exoskeleton: A Plug-and-Play Module to Enhance CNN-Based Occupancy Prediction Networks
Wang, S., Lu, Y., & Ling, Q. (2025) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2025.3595582
“Among image-based methods, [1] and [2] follow detection methods [18], [19] to transform images into 3D space.”
3D Shape Adaptation Across Datasets for Weakly Supervised Monocular 3D Object Detection
Zhang, X., Su, Y., Lu, H., Zhang, C., & Liu, Y. (2025) · IEEE Intelligent Vehicles Symposium · doi:10.1109/iv64158.2025.11097451
3D Shape Transfer Learning for Enhanced Monocular 3D Object Detection
Zhang, X., Su, Y., Lu, H., Zhang, C., Wang, X., & Liu, Y. (2025) · IEEE Intelligent Vehicles Symposium · doi:10.1109/iv64158.2025.11097698
“We prepare the Waymo and nuScenes datasets according to CaDDN [58] and MonoLSS [61], respectively.”
Zeng, C., Wang, Z., Xiao, J., Nguyen, A., Huang, K., Wang, W., & Yue, Y. (2025) · Communications in computer and information science · doi:10.1007/978-981-96-7036-9_22
L4drfusion: Lidar and 4d Millimeter-Wave Radar Fusion in Unified Space for 3d Object DetectionPreprint
Li, Y., Wu, Y., Chen, Z., Wang, M., & Tao, Y. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5383403
Zhang, Z., Li, Z., Wang, H., He, Y., Wang, K., & Fan, H. (2025) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra55743.2025.11128376
HybridBEV: Hybrid Encode and Distillation for Improved BEV 3D Object Detection
Wang, J., Du, C., Liu, H., Xia, Z., Liu, B., & Xiong, S. (2025) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2025.3599015
“For example, Caddn [3] uses predefined point sampling, while LSS [2] relies on depth distribution, both achieving noticeable results.”
InfluentialMGAF: LiDAR-Camera 3D Object Detection With Multiple Guidance and Adaptive Fusion
Fan, B., Li, X., Zhou, Y., Xia, C., Fan, H., Xu, F., & Tian, J. (2025) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2025.3612958
“Furthermore, a large number of works continue to explore Pseudo point clouds 3D detection models, such as Pseudo-LiDAR [14], CaDDN [15] and Pad-net [16].”
Translating Images to Road Network: A Sequence-to-Sequence Perspective
Lu, J., Nie, M., Zhang, B., Peng, R., Cai, X., Xu, H., Wen, F., Zhang, W., & Zhang, L. (2025) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2025.3612940
“To learn ego-car feature learning from onboard cameras, [34], [35] project image features to ego-car coordinate based on depth estimation.”
Unified Self-supervised Representation Learning for Multi-modal and Single-modal 3D Perception
Xu, X., Li, Y., Zhang, T., Yang, J., Johnson‐Roberson, M., & Huang, X. (2025) · IEEE International Conference on Automation Science and Engineering (CASE) · doi:10.1109/case58245.2025.11163793
“1 ) Camera-only models [27], [28] focus on monocular 3D detection [29] and have evolved with larger benchmarks [1], [30] to include range-view images [31] and bird’s-eye view (BEV) perception [16], [32]–[35].”
Semantic-aware testing for object detection systems
Liu, X., Wang, J., Lin, H.-Y., Fang, C., Shi, J., Zhang, X., & Wang, W. (2025) · Information and Software Technology · doi:10.1016/j.infsof.2025.107888
Online 3D Multi-Camera Perception Through Robust 2D Tracking and Depth-Based Late Aggregation
Le, V.-M., Tran, T.-A., Do, D. H., Do, X. C., Ninh, H., & Tran, H. (2025) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw69036.2025.00570
“Depth-distribution models [10, 39] increase precision at the cost of complexity.”
Exploring Vision-Based Active 3D Object Detection by Informativeness Characterization
Li, R., Wu, Y.-M., Lu, Y.-H., Li, X.-W., Wang, X., Liang, X.-B., & Li, X. (2025) · IEEE Transactions on Image Processing · doi:10.1109/tip.2025.3613927
“For vision-based 3D object detection, LSS-style methods [3], [63], [64], [65], [66] and OFT-style methods [66], [67], [68], [69] are two key technologies to achieve 2D-to-3D mapping, which have been proved to be practical for vision-based 3DOD.”
MonoDEA: Monocular 3D Object Detection with Depth-Enhanced Attention
Zhang, X., Ding, Z., Fang, Y., & Ling, Q. (2025) · Chinese Control Conference (CCC) · doi:10.23919/ccc64809.2025.11179658
“M3D-RPN[7] designs depth-aware convolution for better 3D scene understanding, while M3DSSD[8] presents a two-step feature alignment approach to overcome the mismatching problem.”
Yu, Z., & Jin, H. (2025) · Chinese Control Conference (CCC) · doi:10.23919/ccc64809.2025.11179516
“Methods derived from [17,18] project encoded 2D image features into a 3D voxel grid by predicting a discrete depth distribution for each feature point, further generating 3D predictions.”
Ada3DLane: Adaptive 3D Lane Detection From Monocular Images
Hou, Z., Cao, Y., Chen, N., Ren, C., Lin, C., Lang, C., & Li, Y. (2025) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2025.3614189
FidelityBEV: A Fidelity-Preserving Network for Monocular Bird's-Eye-View Semantic Segmentation
Li, S., Chen, P., Zhang, W., Zhang, J., Xing, Y., & Lu, X. (2025) · International Conference on Intelligent Perception and Pattern Recognition (IPPR) · doi:10.1109/ippr66507.2025.11198421
“However, research in BEV perception has predominantly followed a task-splitting paradigm, with one body of work focusing on map element segmentation[4] and another on object detection[5], [6].”
InfluentialMDC: Modality Distribution Consistent Distillation for Multi-View 3D Object Detection
Liu, H., Wang, J., Zhu, W., Fu, B., Xiong, S., & Liu, C. (2025) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme59968.2025.11209314
“Existing methods use depth prediction to complete the image view transformation (IVT) to give 3D geometric information to the image, such as Caddn[7], LSS[8], etc. (green area in Fig.”
Yang, C., Huang, L., Gao, H., Zheng, L., & Chen, F. (2025) · International Conference on Image and Video Processing (ICIVP) · doi:10.1109/icivp66296.2025.00010
“Another approach involves first converting the dense depth map into a pseudo 3D point cloud and then applying laser-based 3D detectors on it to derive results [11][12][13].”
PCG: Enhancing Object Distance Estimation via Patch-to-Center Geometry
Guan, Z., Wang, Z., Hou, J., & Feng, Z. (2025) · IEEE International Conference on Mobile Ad-Hoc and Sensor Systems (MASS) · doi:10.1109/mass66014.2025.00068
Guo, T., & Yu, S. (2025) · International Journal of Automotive Technology · doi:10.1007/s12239-025-00370-1
Erlewein, H., Müller, A. T., Gerstlauer, L. M., Schneider, A. F., & Stache, N. C. (2025) · IEEE International Automated Vehicle Validation Conference (IAVVC) · doi:10.1109/iavvc61942.2025.11219504
CRS3D: Consistency Regularization for Sparsely-supervised 3D object detection
Zeng, B., Wang, Z., Liu, Z., Chen, Y., Liu, W., & Yang, W. (2025) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn64981.2025.11227860
“…3D Object Detection: It is difficult to obtain a high-quality 3D bounding box solely from a single image in monocular camera object detection [20], [21], so binocular or multi-camera detection has been produced, which can infer more accurate depth information through the geometric constraints of…”
Liang, J., Yin, H., Qi, X., Park, J. J., Sun, M., Madhivanan, R., & Manocha, D. (2025) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros60139.2025.11247407
“Early methods [31]–[34] utilized geometric information to correlate 2D image features with 3D space.”
Consistent Feature Alignment for Cross-Modal Knowledge Distillation in Monocular 3D Object Detection
Li, F., Ding, R., Yang, M., & Lan, X. (2025) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros60139.2025.11247216
“Current methods follow four sensor paradigms: LiDAR-based [1]–[3], multi-camera [4]–[6], multi-modal fusion [7]– [9], and monocular methods [10]–[12].”
PLOT: Pseudo-Labeling via Object Tracking for Monocular 3D Object Detection
Lee, S., Aung, S., Choi, J., Kim, S., Kim, I.-J., & Cho, J. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-37235-2_34
Multi-view Point Cloud Object Recognition in Rehabilitation and Daily Life Intelligent Systems
Song, Q., Dou, W., Wang, W., Chyanzheng, S., & Kubota, N. (2025) · Lecture notes in networks and systems · doi:10.1007/978-3-032-10150-1_31
Enhancing collaborative perception through multi-scale contextual information integration
Ma, M., Hong, D., Zhang, H., Miao, Z., Wang, W., & Zhao, G. (2025) · Accident Analysis & Prevention · doi:10.1016/j.aap.2025.108367
BEV-SI: A Lightweight and Efficient Split-Inception Framework for Multi-modal 3D Object Detection
Wu, Y., He, H., & Tang, Y. (2026) · Communications in computer and information science · doi:10.1007/978-981-95-4875-0_13
Wu, G., Li, Q., Jiu, M., Li, S., Zhao, H., & Xu, M. (2025) · IEEE International Conference on Parallel and Distributed Systems (ICPADS) · doi:10.1109/icpads67057.2025.11323154
“Recent advancements attempt to address these issues through multi-frame temporal fusion (e.g., BEVFormer [11]) and physics-aware depth estimation (e.g., CaDDN [12]), but the gap with LiDAR-based systems persists.”
Jiawei, 陆. L., Yangsheng, 钟. Z., Sihan, 葛. G., Jiajun, 叶. Y., & Gang, 肖. X. (2026) · Laser & Optoelectronics Progress · doi:10.3788/lop251539
Chen, W., Chen, M., Fang, J., & zhao, H. (2025) · International Conference on Robotics and Computer Vision (ICRCV) · doi:10.1109/icrcv67407.2025.11349310
RendBEV: Semantic Perspective View Rendering as Supervision for Bird’s Eye View Segmentation
Monteagudo, H. P., Taccari, L., Pjetri, A., Sambo, F., & Salti, S. (2026) · IEEE Access · doi:10.1109/access.2026.3656618
“3D object detection [7], [8], [9], trajectory prediction [10], [11] and accident dynamics reconstruction [12].”
Revisiting Monocular 3D Object Detection With Depth Thickness Field
Zhang, Q., Lin, C., Shen, Z., Nie, L., & Zhao, Y. (2026) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2026.3658133
“Different from existing approaches [30, 33] that adapt Lift-Splat-Shoot (LSS) [31] framework and bird’s-eye-view (BEV) space, MonoDTF retargets traditional depth representations in a new formulation: Depth Thick-ness Field (DTF) to adapt to 3D object detection.”
InfluentialEnd-to-End Monocular Perception and Prediction for Intelligent Driving
Tang, K., & Tian, W. (2025) · CAA International Conference on Vehicular Control and Intelligence (CVCI) · doi:10.1109/cvci66304.2025.11348318
“Feature perspective transform methods are mainly classified into three categories: 1) 2D-3D conversion (e.g., LSS [7], CaDDN [8]) lifts 2D features to 3D space through depth estimation; 2) 3D-2D conversion (e.g., OFT-Net [9], BEVFormer [10]) establishes feature geometric projection based on the IPM…”
Yang, F., Guan, R., Man, K. L., Purwanto, E., Smith, J. S., & Yue, Y. (2026) · IEEE Access · doi:10.1109/access.2026.3660956
“For instance, MonoDepth [27] combines images with depth estimation results as RGBD images to extract 3D features, while LSS [18] and CaDNN [28] directly estimate depth distributions and ‘splat’ image features into BEV space as 3D feature volumes.”
Yoon, J. H., Jung, J. W., & Yoo, S. B. (2026) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2026.3652840
“In contrast, because monocular images provide restricted cues for inferring 3D features, other monocular 3D object detection methods have incorporated supplementary data to enhance detection robustness [29], [30], [31], [32].”
Bidirectional Supervised Fusion for Cross-Modal 3D Object Detection in Autonomous DrivingPreprint
Yue, S., Liu, J., Hao, W., Zhou, S., & cai, Y. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.6205840
Multiview BEV Fusion From Vehicle-on-Board and Roadside Cameras for 3-D Object Detection
Fu, T., Mao, G., Ren, X., Wang, K., Xiang, W., & Ge, X. (2026) · IEEE Sensors Journal · doi:10.1109/jsen.2026.3661270
“…heights and larger bins at higher heights, making it suitable for long-distance detection but overly sparse for short-range objects; and (3) LID [52], which uses bin widths that increase linearly with height, ensuring balanced sampling density and consistent relative error across all height…”
Polarn: Polar BEV Representation with Unified Multi-Modal Fusion for 3D Detection
Chen, J., Gao, F., & Wang, N. (2025) · International Conference on Robotics, Intelligent Control and Artificial Intelligence (RICAI) · doi:10.1109/ricai68060.2025.11384868
Efficient Dual-Branch 3D Occupancy Prediction for Autonomous Driving
Li, Y.-M., Cai, B., & Zhang, Y. (2025) · International Conference on Data Science and Intelligent Systems (DSIS) · doi:10.1109/dsis67228.2025.11390074
“Furthermore, some approaches, such as those proposed in [14] [4], employ virtual cameras to standardize camera parameters, while [7] utilizes Multi-Layer Perceptrons to learn transformation mechanisms from perspective views to bird’s-eye views, among others.”
Kong, H., Xu, L., Peng, L., & An, W. (2025) · International Conference on Frontiers Technology in Circuits and Systems (FTCS) · doi:10.1109/ftcs68006.2025.11405835
Object-Scene-Camera Decomposition and Recomposition for Data Efficient Monocular 3D Object Detection
Kuang, Z., Ding, R., Yang, M., Zheng, X., & Hua, G. (2026) · International Journal of Computer Vision · doi:10.1007/s11263-026-02755-w
SCPM: monocular 3D object detection with spatiotemporal consistent pseudo-labels module
Wang, Y., Rahman, A. H. A., & Rashid, F. A. N. (2026) · Complex & Intelligent Systems · doi:10.1007/s40747-026-02271-x
MonoSC: Enhancing Monocular 3D Object Detection by 2D Segmentation and Completion
Sun, H., Song, Z., Lin, X., & Lu, J. (2025) · IEEE International Conference on Intelligent Transportation Systems (ITSC) · doi:10.1109/itsc60802.2025.11423196
“Among them, CaDDN [12] is a typical method based on depth estimation.”
Improving Monocular 3D Object Detection for Far-Field Scenarios
Gao, Z., Zhao, X., & Chen, H. (2025) · IEEE International Conference on Intelligent Transportation Systems (ITSC) · doi:10.1109/itsc60802.2025.11423705
“CaDDN [11] ranks 1st in depth-based methods on the KITTI 3D detection benchmark, which informatively constructs a new kind of representation using depth estimation and detect objects in BEV (Bird’s Eye View) using the 2D detection method.”
InfluentialDenseRadar: Dense Radar Generation and Point Consistency Aggregation for 3D Object Detection
Wang, J., Xiong, S., Du, C., Li, H., & Xiong, S. (2026) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2026.3678439
“Radar’s inherently sparse perception is complemented by the dense detection capabilities of cameras utilizing LSS [23] and Caddn [36] for a Bird’s-Eye-View (BEV).”
MonoDAPE: Data Augmentation and Perception Enhancement for Robust Monocular 3D Object Detection
Li, Z., Li, Y., Bian, Y., Wang, X., Wang, Y., & Liu, M. (2026) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2026.3678490
“For the Waymo dataset, we follow the same evaluation protocol as [54], [57], using APH 3 D and AP 3 D as metrics and reporting results across the same three distance ranges: [0,30), [30,50), and [50 , ∞ ) meters.”
Target-aware proposal-level fusion for multi-modal three-dimensional detection
Zhao, Z., Wu, B., Yin, Y., Zhang, J., Li, Y., Zeng, Y., & Gao, H. (2026) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2026.114576
SparseBEV: A Fully Sparse Framework for Multi-View 3D Object Detection
Chen, Y., Liu, H., & Wang, L. (2026) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2026.3679808
“Several works [47], [53], [57], [58], [70], [72], [76], [84] incorporate depth information to deal with this problem.”
MVTest: Automated Metamorphic Testing of Multi-view Perception Systems
Luo, L., Pan, Y., & Fan, Y. (2026) · Lecture notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering · doi:10.1007/978-3-032-21171-2_19
Perception Module of Autonomous Driving
Huang, Y., & Yang, Z. (2026) · Springer eBooks · doi:10.1007/978-981-95-3725-9_5
DGKD: Depth-Guided knowledge distillation network for monocular 3D object detection
Zhang, X., & Ling, Q. (2026) · Applied Intelligence · doi:10.1007/s10489-026-07242-w
“Following the partitioning standard of [27], we further divide the training data into a training set with 3712 samples and a validation set with 3769 samples.”
呼木吉力吐 (2025) · China Automation Congress (CAC) · doi:10.1109/cac67268.2025.11487657
“CaDDN [21] provides a fully differentiable end-to-end approach for combining depth estimation and object detection.”
Monocular 3D Object Detection Algorithm Based on Uncertainty Quantification and Few-Shot Learning
Xue, M., Shen, X., Zheng, W., & Li, Z. (2025) · International Conference on Industrial Automation, Robotics and Control Engineering (IARCE) · doi:10.1109/iarce68366.2025.11485676
Unsupervised Domain Adaptation with Multimodal Fusion for Monocular 3D Object Detection
Jiang, J., Dai, J., Li, W., Zhou, Y., Ye, M., Zhang, J.-Y., & Zhang, C. (2026) · Vehicles · doi:10.3390/vehicles8050098
“Subsequent research [12,28,29] continued to advance monocular 3D detection through diverse depth-aware feature representation strategies.”
Wang, C., Huang, Z., Zheng, Z., Hu, Z., & Chen, D. (2026) · Lecture notes in computer science · doi:10.1007/978-981-95-7084-3_13
Lin, Y., & Jia, S. (2026) · Sensors · doi:10.3390/s26113551
OCR-3D: Occlusion-Robust Context-Guided Relative Depth for Monocular 3D Object Detection
Su, J., Hu, C., Xu, Y., Chen, S., Sun, Y., & Su, J. (2026) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2026.3700380
“Moreover,anotherlineofworkrelies ondensedepthsupervisionorauxiliarydepthnetworkstopro-videgeometricpriors[3],[8].”
MonoHPE: Monocular 3D Object Detection With Height Positional Encoding
Zhu, C., Li, Z., Dai, Z., Zheng, L., & Wang, W. (2026) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2026.3700175
FDARC: Frequency-Aware and Depth Association Radar–Camera Fusion
Wang, H., Duan, X., & Zhang, C. (2026) · Electronics · doi:10.3390/electronics15122672
Real-Time 3D Object Detection Algorithm Based on the Fusion of Image and Point Cloud
Shan, W., Chen, X., & Xu, Z. (2026) · Chinese Control and Decision Conference (CCDC) · doi:10.1109/ccdc69976.2026.11559944
“CADDN [1] predicts the depth distribution of pixels by using an optic cone feature network, constructs BEV voxel features and extracts features to predict the 3D location using an existing BEV feature extraction network.”
Yang, T., Qian, Y., Wang, C., & Yang, M. Z. (2026) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2026.3704370
“The forward projection method [9], [30], [36], [37] exemplified by LSS [9], predicts the depth distribution at the pixel level and ultimately derives BEV features through grid sampling, which demonstrates robust geometric perception capabilities.”
MonoPRT: Progressive Relation-Aware Refinement Transformer for Monocular 3D Object Detection
Song, C., Xie, C., & Li, Y. (2026) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2026.3707350
Hu, Z., Liang, B., Wang, Y., Wang, C., Jiang, S., Huang, Z., Hu, H., & Chen, D. (2026) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2026.3710048
“FCOS3D [12] pioneered this by re-formulating the 2D anchor-free FCOS for 3D inference through decoupled geometry, and CaDDN [13] further improved depth estimation by learning pixel-level depth distributions.”
Aljohani, A. A. (2026) · Journal of King Saud University - Computer and Information Sciences · doi:10.1007/s44443-026-01022-9
Mao, G., Fu, T., & Wang, K. (2026) · Transportation Research Record Journal of the Transportation Research Board · doi:10.1177/03611981261445067
Duan, Z., Hu, X., An, P., Dang, C., Zhang, X., & Duan, Z. (2026) · Pattern Recognition · doi:10.1016/j.patcog.2026.114453
Christanto, R., & Miaou, S.-G. (2026) · Sensors · doi:10.3390/s26144440
“However, these models inherently rely on dense geometric feature extraction [9,10] and large model capacity, limiting their deployment on edge devices.”
FlashBEV: Fast and Memory-Efficient Exact BEV Transformation with IO-Awareness
Yokokawa, S., & Kasahara, H. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-37362-5_29
Yang, J., Zhang, Y., Zhang, Y., & Zhang, Y. (2026) · Sensors · doi:10.3390/s26144615
“Furthermore, CaDDN [16] incorporates geometric reasoning through depth distribution estimation, while uncertainty-aware ev-idential fusion frameworks [17] explicitly model sensor uncertainty.”
Sekaran, K. C., Kalyanasundaram, A., Botsch, M., & Utschick, W. (2026) · IEEE Intelligent Vehicles Symposium · doi:10.1109/iv66570.2026.11623881
“Depth-aware approaches first estimate per-pixel depth and then fuse it with image features to improve distance estimation [21], [22], [23], [24], while pseudo-LiDAR methods lift the predicted depth into 3D space to form point clouds processed by LiDAR-style detectors [25].”
Liu, K., Ma, X., Hu, R., Liu, B., Wang, Z., & Wang, Z. (2026) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2026.3724359
“We mainly adopt CaDDN [9], MonoDTR [4], DID-M3D [41], Cube R-CNN [42], MonoUNI [43], MVC-MonoDet [28], CMKD [37], Mix-Teaching [13], DPL [29] and OccupancyM3D [12] as baselines and mainly evaluate the results of the Car category.”
InfluentialGeoSemi-Bike: Dual-Branch Geometric-Semantic Reasoning for E-Bike Governance
Xu, K., Yue, K., & Hei, X. (2026) · Information Communication Technologies Conference (ICTC) · doi:10.1109/ictc70246.2026.11650179
“For dockless e-bikes in open scenarios, despite monocular 3D object detection progress [5], [6], applying general models for specific e-bike geometric constraints faces enormous recognition challenges in long-tail scenarios.”
Li, J., Li, G., Liu, Q., Xiong, S., Zhao, Z., & Wang, H. (2026) · Journal of King Saud University - Computer and Information Sciences · doi:10.1007/s44443-026-01251-y
“CaDDN (Reading et al. 2021) introduces category-aware depth-distribution modeling.”
Segmentation Residual–Gated Multi-Scale Region Enhancement in Monocular 3D Object Detection
Peng, T., An, J., & Kim, B.-W. (2026) · Transactions of Korean Society of Automotive Engineers · doi:10.7467/ksae.2026.34.9.1009
MonoTemp3D: Lightweight Short-Horizon Temporal Enhancement for Streaming Monocular 3D Car Detection
Li, J., Liu, T., Guo, X., & Li, Y. (2026) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2026.3732838
“Depth-assisted methods utilize predicted dense depth maps [11] or categorical depth distributions [12] to construct stronger 3D representations, or incorporate depth cues into structured feature propagation [23].”
Tang, J., Hou, Y., Bo, Y., & Liu, S. (2026) · Journal of Field Robotics · doi:10.1002/rob.70348
Chen, T., Cui, H., Ke, Z., Han, Y., & Zhang, G. (2026) · The Journal of Supercomputing · doi:10.1007/s11227-026-08874-y
MonoKey: Occlusion-Robust Keypoint-based Monocular 3D Object Detection with Prior Guidance
Cho, Y. W., Cheon, J. W., Yoon, J. H., & Yoo, S. B. (2026) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra57385.2026.11697155
“In response to the challenges of monocular perception, some methods have incorporated auxiliary signals, such as generating bird’s-eye view representations or using 2D corner projections [19], [20].”
Zhang, K., As’arry, A., Lian, S., Hairuddin, A. A., Hassan, M. K., Qin, W., & Wei, J. (2026) · Journal of Materials Engineering and Performance · doi:10.1007/s11665-026-15231-w
D2OM: Depth-aware dynamic matching for DETR-based monocular 3D detection
Wei, C., Zuo, S., Li, C., Yang, M., Qi, Y., Xie, G., Wang, S., & Yang, Z. (2026) · Knowledge-Based Systems · doi:10.1016/j.knosys.2026.117140
Liu, Y., Guo, L., Zhang, Q., Duan, C., Zhang, X., Lei, Z., Chen, X., & Peng, X. (2026) · Electronics · doi:10.3390/electronics15204821
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
