{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,4]],"date-time":"2026-02-04T01:00:39Z","timestamp":1770166839177,"version":"3.49.0"},"reference-count":50,"publisher":"MDPI AG","issue":"3","license":[{"start":{"date-parts":[[2026,2,2]],"date-time":"2026-02-02T00:00:00Z","timestamp":1769990400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Sensors"],"abstract":"<jats:p>This paper proposes a novel motion-aware framework to address key challenges in 4D millimeter-wave radar detection for autonomous driving. While existing methods struggle with sparse point clouds and dynamic object characterization, our approach introduces three key innovations: (1) A Bird\u2019s Eye View (BEV) fusion network incorporating velocity vector decomposition and dynamic gating mechanisms, effectively encoding motion patterns through independent XY-component convolutions; (2) a gradient-aware multi-task balancing scheme using learnable uncertainty parameters and dynamic weight normalization, resolving optimization conflicts between classification and regression tasks; and (3) a two-phase progressive training strategy combining multi-frame pre-training with sparse single-frame refinement. Evaluated on the TJ4D benchmark, our method achieves 33.25% mean Average Precision (mAP)3D with minimal parameter overhead (1.73 M), showing particular superiority in pedestrian detection (+4.16% AP) while maintaining real-time performance (24.4 FPS on embedded platforms). Comprehensive ablation studies validate each component\u2019s contribution, with thermal map visualization demonstrating effective motion pattern learning. This work advances robust perception under challenging conditions through principled motion modeling and efficient architecture design.<\/jats:p>","DOI":"10.3390\/s26030968","type":"journal-article","created":{"date-parts":[[2026,2,3]],"date-time":"2026-02-03T12:46:11Z","timestamp":1770122771000},"page":"968","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Dynamic Feature Fusion for Sparse Radar Detection: Motion-Centric BEV Learning with Adaptive Task Balancing"],"prefix":"10.3390","volume":"26","author":[{"given":"Yixun","family":"Sang","sequence":"first","affiliation":[{"name":"School of Mechanical and Electrical Engineering, North University of China, Taiyuan 030051, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Junjie","family":"Cui","sequence":"additional","affiliation":[{"name":"School of Mechanical and Electrical Engineering, North University of China, Taiyuan 030051, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yaoguang","family":"Sun","sequence":"additional","affiliation":[{"name":"School of Mechanical and Electrical Engineering, North University of China, Taiyuan 030051, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Fan","family":"Zhang","sequence":"additional","affiliation":[{"name":"School of Mechanical and Electrical Engineering, North University of China, Taiyuan 030051, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yanting","family":"Li","sequence":"additional","affiliation":[{"name":"School of Mechanical and Electrical Engineering, North University of China, Taiyuan 030051, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Guoqiang","family":"Shi","sequence":"additional","affiliation":[{"name":"School of Mechanical and Electrical Engineering, North University of China, Taiyuan 030051, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2026,2,2]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"21780","DOI":"10.1109\/JSEN.2022.3197235","article-title":"Pseudo RGB-D Face Recognition","volume":"22","author":"Jin","year":"2022","journal-title":"IEEE Sens. J."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"123881","DOI":"10.1016\/j.eswa.2024.123881","article-title":"Simulated multimodal deep facial diagnosis","volume":"252","author":"Jin","year":"2024","journal-title":"Expert Syst. Appl."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"5575","DOI":"10.1109\/TAI.2024.3421175","article-title":"Social NSTransformers: Low-Quality Pedestrian Trajectory Prediction","volume":"5","author":"Jiang","year":"2024","journal-title":"IEEE Trans. Artif. Intell."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"16438","DOI":"10.1109\/TITS.2025.3572254","article-title":"Social Entropy Informer: A Multi-Scale Model-Data Dual-Driven Approach for Pedestrian Trajectory Prediction","volume":"26","author":"Jiang","year":"2025","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"ref_5","first-page":"15","article-title":"Social Informer: Pedestrian Trajectory Prediction by Informer with Adaptive Trajectory Probability Region Optimization","volume":"26","author":"Jiang","year":"2025","journal-title":"IEEE Trans. Cybern."},{"key":"ref_6","doi-asserted-by":"crossref","unstructured":"Yan, Y., Mao, Y., and Li, B. (2018). SECOND: Sparsely Embedded Convolutional Detection. Sensors, 18.","DOI":"10.3390\/s18103337"},{"key":"ref_7","doi-asserted-by":"crossref","unstructured":"Lang, A.H., Vora, S., Caesar, H., Zhou, L., Yang, J., and Beijbom, O. (2019, January 15\u201320). PointPillars: Fast Encoders for Object Detection From Point Clouds. Proceedings of the 2019 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Long Beach, CA, USA.","DOI":"10.1109\/CVPR.2019.01298"},{"key":"ref_8","doi-asserted-by":"crossref","unstructured":"Yin, T., Zhou, X., and Kr\u00e4henb\u00fchl, P. (2021, January 20\u201325). Center-based 3D Object Detection and Tracking. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, TN, USA.","DOI":"10.1109\/CVPR46437.2021.01161"},{"key":"ref_9","doi-asserted-by":"crossref","unstructured":"Li, J., Luo, C., and Yang, X. (2023, January 17\u201324). PillarNeXt: Rethinking Network Designs for 3D Object Detection in LiDAR Point Clouds. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Vancouver, BC, Canada.","DOI":"10.1109\/CVPR52729.2023.01685"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"799","DOI":"10.1109\/TIV.2023.3322729","article-title":"SMURF: Spatial Multi-Representation Fusion for 3D Object Detection With 4D Imaging Radar","volume":"9","author":"Liu","year":"2024","journal-title":"IEEE Trans. Intell. Veh."},{"key":"ref_11","doi-asserted-by":"crossref","unstructured":"Xiong, W., Liu, J., Huang, T., Han, Q.L., Xia, Y., and Zhu, B. (2024, January 2\u20135). LXL: LiDAR Excluded Lean 3D Object Detection with 4D Imaging Radar and Camera Fusion. Proceedings of the 2024 IEEE Intelligent Vehicles Symposium (IV), Jeju Island, Republic of Korea.","DOI":"10.1109\/IV55156.2024.10588781"},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"8641","DOI":"10.1109\/TITS.2025.3554313","article-title":"MSSF: A 4D Radar and Camera Fusion Framework With Multi-Stage Sampling for 3D Object Detection in Autonomous Driving","volume":"26","author":"Liu","year":"2025","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Peng, X., Tang, M., Sun, H., Bierzynski, K., Servadei, L., and Wille, R. (2024, January 17\u201320). MUFASA: Multi-view Fusion and Adaptation Network with Spatial Awareness for Radar Object Detection. Proceedings of the Artificial Neural Networks and Machine Learning\u2014ICANN 2024, Lugano, Switzerland.","DOI":"10.1007\/978-3-031-72335-3_12"},{"key":"ref_14","doi-asserted-by":"crossref","unstructured":"Jia, L., Guan, R., Zhao, H., Zhao, Q., Man, K.L., Smith, J.S., Yu, L., and Yue, Y. (2025). RadarNeXt: Real-Time and Reliable 3D Object Detector Based On 4D mmWave Imaging Radar. arXiv.","DOI":"10.1186\/s13634-025-01271-2"},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"9981","DOI":"10.1109\/TITS.2021.3096943","article-title":"Labels Are Not Perfect: Inferring Spatial Uncertainty in Object Detection","volume":"23","author":"Feng","year":"2022","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Mandalika, S., V, L., and Nambiar, A. (2025, January 11\u201312). PRIMEDrive-CoT: A Precognitive Chain-of-Thought Framework for Uncertainty-Aware Object Interaction in Driving Scene Scenario. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), Nashville, TN, USA.","DOI":"10.1109\/CVPRW67362.2025.00526"},{"key":"ref_17","doi-asserted-by":"crossref","unstructured":"Peng, X., Liu, S., Gao, C., Bai, Y., Mu, B., Wang, X., and Xia, H. (2025). GLRD: Global-Local Collaborative Reason and Debate with PSL for 3D Open-Vocabulary Detection. arXiv.","DOI":"10.1007\/978-3-031-72761-0_21"},{"key":"ref_18","doi-asserted-by":"crossref","unstructured":"He, C., Li, R., Zhang, Y., Li, S., and Zhang, L. (2023, January 18\u201322). MSF: Motion-guided Sequential Fusion for Efficient 3D Object Detection from Point Cloud Sequences. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Vancouver, BC, Canada.","DOI":"10.1109\/CVPR52729.2023.00503"},{"key":"ref_19","doi-asserted-by":"crossref","unstructured":"McCrae, S., and Zakhor, A. (2020, January 25\u201328). 3d Object Detection For Autonomous Driving Using Temporal Lidar Data. Proceedings of the 2020 IEEE International Conference on Image Processing (ICIP), Abu Dhabi, United Arab Emirates.","DOI":"10.1109\/ICIP40778.2020.9191134"},{"key":"ref_20","doi-asserted-by":"crossref","unstructured":"Bharadhwaj, B.M., and Nair, B.B. (2022, January 24\u201326). Deep Learning-based 3D Object Detection Using LiDAR and Image Data Fusion. Proceedings of the 2022 IEEE 19th India Council International Conference (INDICON), Kochi, India.","DOI":"10.1109\/INDICON56171.2022.10040030"},{"key":"ref_21","doi-asserted-by":"crossref","unstructured":"Khan, D., Baek, M., Kim, M.Y., and Seog Han, D. (2022, January 19\u201321). Multimodal Object Detection and Ranging Based on Camera and Lidar Sensor Fusion for Autonomous Driving. Proceedings of the 2022 27th Asia Pacific Conference on Communications (APCC), Jeju Island, Republic of Korea.","DOI":"10.1109\/APCC55198.2022.9943618"},{"key":"ref_22","doi-asserted-by":"crossref","unstructured":"Wu, H., Wen, C., Shi, S., and Wang, C. (2023, January 17\u201324). Virtual Sparse Convolution for Multimodal 3D Object Detection. Proceedings of the CVPR, Vancouver, BC, Canada.","DOI":"10.1109\/CVPR52729.2023.02074"},{"key":"ref_23","unstructured":"Son, H., He, J., Park, S.I., Min, Y., Zhang, Y., and Yoo, B. (2025). SparseVoxFormer: Sparse Voxel-based Transformer for Multi-modal 3D Object Detection. arXiv."},{"key":"ref_24","doi-asserted-by":"crossref","unstructured":"Zhang, B., Yuan, J., Shi, B., Chen, T., Li, Y., and Qiao, Y. (2023, January 17\u201324). Uni3D: A Unified Baseline for Multi-dataset 3D Object Detection. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Vancouver, BC, Canada.","DOI":"10.1109\/CVPR52729.2023.00893"},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"37175","DOI":"10.1109\/JIOT.2025.3582636","article-title":"RailVoxelDet: A Lightweight 3-D Object Detection Method for Railway Transportation Driven by Onboard LiDAR Data","volume":"12","author":"Chen","year":"2025","journal-title":"IEEE Internet Things J."},{"key":"ref_26","doi-asserted-by":"crossref","unstructured":"Cheng, J.H., Kuan, S.Y., Liu, H.I., Latapie, H., Liu, G., and Hwang, J.N. (2024, January 27\u201330). CenterRadarNet: Joint 3D Object Detection and Tracking Framework Using 4D FMCW Radar. Proceedings of the 2024 IEEE International Conference on Image Processing (ICIP), Abu Dhabi, United Arab Emirates.","DOI":"10.1109\/ICIP51287.2024.10648077"},{"key":"ref_27","doi-asserted-by":"crossref","unstructured":"Xu, B., Zhang, X., Wang, L., Hu, X., Li, Z., Pan, S., Li, J., and Deng, Y. (2021, January 19\u201322). RPFA-Net: A 4D RaDAR Pillar Feature Attention Network for 3D Object Detection. Proceedings of the 2021 IEEE International Intelligent Transportation Systems Conference (ITSC), Indianapolis, IN, USA.","DOI":"10.1109\/ITSC48978.2021.9564754"},{"key":"ref_28","doi-asserted-by":"crossref","unstructured":"Zou, Y., Li, J., Chen, L., and Yang, B. (2024, January 19\u201321). VA-Net: 3D Object Detection with 4D Radar Based on Self-Attention. Proceedings of the International Conference on Computer Vision and Deep Learning, Changsha, China.","DOI":"10.1145\/3653804.3654611"},{"key":"ref_29","doi-asserted-by":"crossref","first-page":"5102","DOI":"10.1109\/JSEN.2024.3516786","article-title":"Multiscale Pillars Fusion for 4-D Radar Object Detection With Radar Data Enhancement","volume":"25","author":"Wang","year":"2025","journal-title":"IEEE Sens. J."},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"4961","DOI":"10.1109\/LRA.2022.3147324","article-title":"Multi-Class Road User Detection With 3+1D Radar in the View-of-Delft Dataset","volume":"7","author":"Palffy","year":"2022","journal-title":"IEEE Robot. Autom. Lett."},{"key":"ref_31","doi-asserted-by":"crossref","unstructured":"Dam, T., Dharavath, S.B., Alam, S., Lilith, N., Maiti, A., Chakraborty, S., and Feroskhan, M. (2025, January 19\u201323). SaViD: Spectravista Aesthetic Vision Integration for Robust and Discerning 3D Object Detection in Challenging Environments. Proceedings of the 2025 IEEE International Conference on Robotics and Automation (ICRA), Atlanta, GA, USA.","DOI":"10.1109\/ICRA55743.2025.11128252"},{"key":"ref_32","doi-asserted-by":"crossref","unstructured":"Hu, X., Duan, Z., An, P., Zhang, J., and Ma, J. (2025). Dual-Domain Homogeneous Fusion with Cross-Modal Mamba and Progressive Decoder for 3D Object Detection. arXiv.","DOI":"10.1016\/j.patcog.2025.112725"},{"key":"ref_33","doi-asserted-by":"crossref","unstructured":"Mali\u0107, D., Fruhwirth-Reisinger, C., Schulter, S., and Possegger, H. (2025, January 10\u201317). GBlobs: Explicit Local Structure via Gaussian Blobs for Improved Cross-Domain LiDAR-based 3D Object Detection. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, TN, USA.","DOI":"10.1109\/CVPR52734.2025.02548"},{"key":"ref_34","doi-asserted-by":"crossref","unstructured":"Rong, Y., Wei, X., Lin, T., Wang, Y., and Kasneci, E. (2023, January 17\u201324). DynStatF: An Efficient Feature Fusion Strategy for LiDAR 3D Object Detection. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), Vancouver, BC, Canada.","DOI":"10.1109\/CVPRW59228.2023.00326"},{"key":"ref_35","unstructured":"Zhang, R., Guan, R., Chen, X., Prugel-Bennett, A., and Cai, X. (2025). CornerPoint3D: Look at the Nearest Corner Instead of the Center. arXiv."},{"key":"ref_36","unstructured":"Bheemaiah, A., and Yang, S. (2025). Knowledge Graphs as World Models for Semantic Material-Aware Obstacle Handling in Autonomous Vehicles. arXiv."},{"key":"ref_37","doi-asserted-by":"crossref","unstructured":"Ra, S.T., and Lee, S.H. (2025, January 19\u201322). Max Intensity Voxelization Method for Autonomous Driving 3D Object Detection Based on LiDAR. Proceedings of the 2025 International Conference on Electronics, Information, and Communication (ICEIC), Osaka, Japan.","DOI":"10.1109\/ICEIC64972.2025.10879701"},{"key":"ref_38","unstructured":"You, K., Gu, Y., Lin, Y., and Wang, Y. (2025). A novel physical constraint-guided quadratic neural networks for interpretable bearing fault diagnosis under zero-fault sample. Nondestruct. Test. Eval."},{"key":"ref_39","doi-asserted-by":"crossref","unstructured":"Li, J., Yang, L., Chen, Y., Yang, Y., Jin, Y., and Akiyama, K. (2023, January 24\u201328). PillarDAN: Pillar-based Dual Attention Attention Network for 3D Object Detection with 4D RaDAR. Proceedings of the 2023 IEEE 26th International Conference on Intelligent Transportation Systems (ITSC), Bilbao, Spain.","DOI":"10.1109\/ITSC57777.2023.10422406"},{"key":"ref_40","doi-asserted-by":"crossref","unstructured":"Musiat, A., Reichardt, L., Schulze, M., and Wasenm\u00fcller, O. (2024, January 24\u201327). RadarPillars: Efficient Object Detection from 4D Radar Point Clouds. Proceedings of the 2024 IEEE Intelligent Transportation Systems Conference (ITSC), Edmonton, AB, Canada.","DOI":"10.1109\/ITSC58415.2024.10919920"},{"key":"ref_41","doi-asserted-by":"crossref","first-page":"4284","DOI":"10.1109\/LRA.2025.3550707","article-title":"MAFF-Net: Enhancing 3D Object Detection With 4D Radar via Multi-Assist Feature Fusion","volume":"10","author":"Bi","year":"2025","journal-title":"IEEE Robot. Autom. Lett."},{"key":"ref_42","doi-asserted-by":"crossref","first-page":"11125","DOI":"10.1109\/JSEN.2022.3219643","article-title":"3-D Object Detection for Multiframe 4-D Automotive Millimeter-Wave Radar Point Cloud","volume":"23","author":"Tan","year":"2023","journal-title":"IEEE Sens. J."},{"key":"ref_43","doi-asserted-by":"crossref","first-page":"10272","DOI":"10.1109\/JSEN.2025.3528986","article-title":"DADAN: Dynamic-Augmented and Density-Aware Network for Accurate 3-D Object Detection With 4-D Radar","volume":"25","author":"Wang","year":"2025","journal-title":"IEEE Sens. J."},{"key":"ref_44","doi-asserted-by":"crossref","unstructured":"Pan, Z., Ding, F., Zhong, H., and Lu, C.X. (2024, January 13\u201317). RaTrack: Moving Object Detection and Tracking with 4D Radar Point Cloud. Proceedings of the 2024 IEEE International Conference on Robotics and Automation (ICRA), Yokohama, Japan.","DOI":"10.1109\/ICRA57147.2024.10610368"},{"key":"ref_45","doi-asserted-by":"crossref","unstructured":"Chae, Y., Kim, H., and Yoon, K.J. (2024, January 16\u201322). Towards Robust 3D Object Detection with LiDAR and 4D Radar Fusion in Various Weather Conditions. Proceedings of the 2024 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA.","DOI":"10.1109\/CVPR52733.2024.01436"},{"key":"ref_46","doi-asserted-by":"crossref","unstructured":"Xu, R., Xiang, Z., Zhang, C., Zhong, H., Zhao, X., Dang, R., Xu, P., Pu, T., and Liu, E. (March, January 25). SCKD: Semi-Supervised Cross-Modality Knowledge Distillation for 4D Radar Object Detection. Proceedings of the 39th AAAI Conference on Artificial Intelligence (AAAI), Philadelphia, PA, USA.","DOI":"10.1609\/aaai.v39i9.32966"},{"key":"ref_47","doi-asserted-by":"crossref","unstructured":"Zeller, M., Sandhu, V.S., Mersch, B., Behley, J., Heidingsfeld, M., and Stachniss, C. (June, January 29). Radar Velocity Transformer: Single-scan Moving Object Segmentation in Noisy Radar Point Clouds. Proceedings of the 2023 IEEE International Conference on Robotics and Automation (ICRA), London, UK.","DOI":"10.1109\/ICRA48891.2023.10161152"},{"key":"ref_48","doi-asserted-by":"crossref","unstructured":"Guo, Y., Peng, G., and Gao, Q. (2025, January 19\u201323). RadarMask: A Novel End-to-End Sparse Millimeter-Wave Radar Sequence Panoptic Segmentation and Tracking Method. Proceedings of the 2025 IEEE International Conference on Robotics and Automation (ICRA), Atlanta, GA, USA.","DOI":"10.1109\/ICRA55743.2025.11128555"},{"key":"ref_49","doi-asserted-by":"crossref","unstructured":"Paek, D.H., and Kong, S.H. (2025, January 22\u201325). SpikingRTNH: Spiking Neural Network for 4D Radar Object Detection. Proceedings of the 2025 IEEE Intelligent Vehicles Symposium (IV), Cluj-Napoca, Romania.","DOI":"10.1109\/IV64158.2025.11097403"},{"key":"ref_50","doi-asserted-by":"crossref","unstructured":"Zhou, Y., Hao, J., and Zhu, K. (2023, January 27\u201329). RMSA-Net: A 4D Radar Based Multi-Scale Attention Network for 3D Object Detection. Proceedings of the 2023 7th International Symposium on Computer Science and Intelligent Control (ISCSIC), Nanjing, China.","DOI":"10.1109\/ISCSIC60498.2023.00047"}],"container-title":["Sensors"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1424-8220\/26\/3\/968\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,2,3]],"date-time":"2026-02-03T12:51:29Z","timestamp":1770123089000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1424-8220\/26\/3\/968"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,2,2]]},"references-count":50,"journal-issue":{"issue":"3","published-online":{"date-parts":[[2026,2]]}},"alternative-id":["s26030968"],"URL":"https:\/\/doi.org\/10.3390\/s26030968","relation":{},"ISSN":["1424-8220"],"issn-type":[{"value":"1424-8220","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,2,2]]}}}