{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,2]],"date-time":"2026-10-02T06:34:13Z","timestamp":1790922853372,"version":"4.1.0"},"reference-count":37,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2025,11,20]],"date-time":"2025-11-20T00:00:00Z","timestamp":1763596800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2025,11,20]],"date-time":"2025-11-20T00:00:00Z","timestamp":1763596800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Sci Rep"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>This paper addresses perception degradation caused by adverse weather, occlusion, and asynchronous sampling by proposing an uncertainty-weighted multi-task learning framework for robust semantic understanding of traffic scenes (UW-MTL). The method performs differentiable multi-source spatiotemporal alignment to unify camera, LiDAR, radar, and IMU into a BEV sequence, and adopts a hybrid backbone that combines a Mixture of Experts Transformer with a spatiotemporal graph neural network to balance global semantics and local topology. Each task employs evidential prediction heads that explicitly output confidence and uncertainty. During training, soft-temperature weighting and a sigma aware gradient conflict resolver enable stable joint optimization. On the nuScenes benchmark, UW-MTL consistently surpasses BEVFusion and UniAD on 3D object detection, BEV semantic segmentation, and short-horizon trajectory prediction, with especially pronounced gains at long range, under heavy occlusion, and in low-visibility conditions.<\/jats:p>","DOI":"10.1038\/s41598-025-24838-3","type":"journal-article","created":{"date-parts":[[2025,11,20]],"date-time":"2025-11-20T10:42:54Z","timestamp":1763635374000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":4,"title":["Uncertainty weighted multi task learning for robust traffic scene semantic understanding"],"prefix":"10.1038","volume":"15","author":[{"given":"Zhiping","family":"Wan","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shitong","family":"Ye","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Feng","family":"Wang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shaojiang","family":"Liu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ling","family":"Peng","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2025,11,20]]},"reference":[{"key":"24838_CR1","doi-asserted-by":"publisher","first-page":"1291875","DOI":"10.3389\/fnbot.2023.1291875","volume":"17","author":"M Ye","year":"2023","unstructured":"Ye, M. & Zhang, J. Mobip: a lightweight model for driving perception using MobileNet[J]. Front. Neurorobotics. 17, 1291875 (2023).","journal-title":"Front. Neurorobotics."},{"key":"24838_CR2","unstructured":"Wang, J. et al. A survey on deep multi-task learning in connected autonomous vehicles[J]. arXiv preprint arXiv:2508.00917 (2025)."},{"issue":"15","key":"24838_CR3","doi-asserted-by":"publisher","first-page":"3063","DOI":"10.3390\/electronics13153063","volume":"13","author":"N Aloufi","year":"2024","unstructured":"Aloufi, N., Alnori, A. & Basuhail, A. Enhancing autonomous vehicle perception in adverse weather: A multi objectives model for integrated weather classification and object detection[J]. Electronics 13 (15), 3063 (2024).","journal-title":"Electronics"},{"key":"24838_CR4","unstructured":"Huang, T. et al. V2X cooperative perception for autonomous driving: recent advances and challenges[J]. (2023). arXiv preprint arXiv:2310.03525."},{"key":"24838_CR5","unstructured":"Xie, E. et al. M $^ 2$ BEV: multi-camera joint 3D detection and segmentation with unified birds-eye view representation[J]. (2022). arXiv preprint arXiv:2204.05088."},{"key":"24838_CR6","doi-asserted-by":"crossref","unstructured":"Li, Z. et al. Fb-bev: Bev representation from forward-backward view transformations[C]\/\/Proceedings of the IEEE\/CVF International Conference on Computer Vision. : 6919\u20136928. (2023).","DOI":"10.1109\/ICCV51070.2023.00637"},{"key":"24838_CR7","doi-asserted-by":"crossref","unstructured":"Chen, X. et al. M3net: Multimodal multi-task learning for 3d detection, segmentation, and occupancy prediction in autonomous driving[C]\/\/Proceedings of the AAAI Conference on Artificial Intelligence. 39(2): 2275\u20132283. (2025).","DOI":"10.1609\/aaai.v39i2.32227"},{"key":"24838_CR8","doi-asserted-by":"crossref","unstructured":"Luu, V. T. et al. RC-AutoCalib: An end-to-end radar-camera automatic calibration network[C]\/\/proceedings of the computer vision and pattern recognition conference. : 6700\u20136709. (2025).","DOI":"10.1109\/CVPR52734.2025.00628"},{"key":"24838_CR9","doi-asserted-by":"publisher","first-page":"123249","DOI":"10.1016\/j.eswa.2024.123249","volume":"246","author":"M Usman","year":"2024","unstructured":"Usman, M., Zaka-Ud-Din, M. & Ling, Q. Enhanced encoder\u2013decoder architecture for visual perception multitasking of autonomous driving[J]. Expert Syst. Appl. 246, 123249 (2024).","journal-title":"Expert Syst. Appl."},{"key":"24838_CR10","doi-asserted-by":"crossref","unstructured":"Xia, Z. X. et al. Robust long-range perception against sensor misalignment in autonomous vehicles[C]\/\/2025 IEEE\/CVF winter conference on applications of computer vision (WACV). IEEE, : 5761\u20135770. (2025).","DOI":"10.1109\/WACV61041.2025.00562"},{"key":"24838_CR11","doi-asserted-by":"crossref","unstructured":"Vora, S. et al. Pointpainting: Sequential fusion for 3d object detection[C]\/\/Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition. : 4604\u20134612. (2020).","DOI":"10.1109\/CVPR42600.2020.00466"},{"key":"24838_CR12","doi-asserted-by":"crossref","unstructured":"Liu, Z., Tang, H., Amini, A. et al. Bevfusion: Multi-task multi-sensor fusion with unified bird\u2019s-eye view representation[J]. (2022). IEEEInternational Conference on Robotics and Automation (ICRA). IEEE, 2023: 2774\u20132781. arXiv preprint arXiv:2205.13542.","DOI":"10.1109\/ICRA48891.2023.10160968"},{"key":"24838_CR13","doi-asserted-by":"crossref","unstructured":"Bai, X. et al. Transfusion: Robust lidar-camera fusion for 3d object detection with transformers[C]\/\/Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition. : 1090\u20131099. (2022).","DOI":"10.1109\/CVPR52688.2022.00116"},{"key":"24838_CR14","doi-asserted-by":"crossref","unstructured":"Philion, J. & Fidler, S. Lift, splat, shoot: Encoding images from arbitrary camera rigs by implicitly unprojecting to 3d[C]\/\/European conference on computer vision. Cham: Springer International Publishing, : 194\u2013210. (2020).","DOI":"10.1007\/978-3-030-58568-6_12"},{"key":"24838_CR15","doi-asserted-by":"crossref","unstructured":"Li, Z. et al. Bevformer: Learning bird\u2019s-eye-view representation from multi-camera images via spatiotemporal transformers.(2022)[J]. URL (2022). https:\/\/arxiv.org\/abs\/2203.17270.","DOI":"10.1007\/978-3-031-20077-9_1"},{"key":"24838_CR16","unstructured":"Zhang, Y. et al. BEVerse: unified perception and prediction in birds-eye-view for vision-centric autonomous driving[J]. (2022). arXiv preprint arXiv:2205.09743."},{"key":"24838_CR17","doi-asserted-by":"crossref","unstructured":"Hu, Y. et al. Planning-oriented autonomous driving[C]\/\/Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition. : 17853\u201317862. (2023).","DOI":"10.1109\/CVPR52729.2023.01712"},{"key":"24838_CR18","doi-asserted-by":"crossref","unstructured":"Gao, J. et al. Vectornet: Encoding hd maps and agent dynamics from vectorized representation[C]\/\/Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition. : 11525\u201311533. (2020).","DOI":"10.1109\/CVPR42600.2020.01154"},{"key":"24838_CR19","doi-asserted-by":"crossref","unstructured":"Liang, M. et al. Learning lane graph representations for motion forecasting[C]\/\/European Conference on Computer Vision. Cham: Springer International Publishing, : 541\u2013556. (2020).","DOI":"10.1007\/978-3-030-58536-5_32"},{"key":"24838_CR20","doi-asserted-by":"crossref","unstructured":"Kendall, A., Gal, Y. & Cipolla, R. Multi-task learning using uncertainty to weigh losses for scene geometry and semantics[C]\/\/Proceedings of the IEEE conference on computer vision and pattern recognition. : 7482\u20137491. (2018).","DOI":"10.1109\/CVPR.2018.00781"},{"key":"24838_CR21","unstructured":"Sensoy, M., Kaplan, L. & Kandemir, M. Evidential deep learning to quantify classification uncertainty. Advances in NeuralInformation Processing Systems: NeurIPS, 3179\u20133189. (2018)."},{"key":"24838_CR22","doi-asserted-by":"crossref","unstructured":"Yang, Y. et al. Multi-task dense prediction via mixture of low-rank experts[C]\/\/Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition. : 27927\u201327937. (2024).","DOI":"10.1109\/CVPR52733.2024.02638"},{"key":"24838_CR23","doi-asserted-by":"crossref","unstructured":"Zhu, P. et al. Task-customized mixture of adapters for general image fusion[C]\/\/Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition. : 7099\u20137108. (2024).","DOI":"10.1109\/CVPR52733.2024.00678"},{"key":"24838_CR24","unstructured":"Liu, T. et al. Routers in vision mixture of experts: an empirical study[J]. (2024). arXiv preprint arXiv:2401.15969."},{"key":"24838_CR25","doi-asserted-by":"crossref","unstructured":"Jin, M. et al. A Survey on Graph Neural Networks for time Series: Forecasting, classification, imputation, and Anomaly detection[J] (IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024).","DOI":"10.1109\/TPAMI.2024.3443141"},{"key":"24838_CR26","unstructured":"Li, H. et al. A survey on graph neural networks in intelligent transportation systems[J]. (2024). arXiv preprint arXiv:2401.00713."},{"key":"24838_CR27","doi-asserted-by":"crossref","unstructured":"Samiuddin, J., Boulet, B. & Wu, D. Trajectory prediction for autonomous driving using agent-interaction graph embedding[C]\/\/2024 IEEE 27th International Conference on Intelligent Transportation Systems (ITSC). IEEE, : 2259\u20132265. (2024).","DOI":"10.1109\/ITSC58415.2024.10920250"},{"key":"24838_CR28","doi-asserted-by":"crossref","unstructured":"Chen, T. et al. Adamv-moe: Adaptive multi-task vision mixture-of-experts[C]\/\/proceedings of the IEEE\/CVF international conference on computer vision. : 17346\u201317357. (2023).","DOI":"10.1109\/ICCV51070.2023.01591"},{"key":"24838_CR29","unstructured":"Liu, W. et al. TEM^ 3-learning: time-efficient multimodal multi-task learning for advanced assistive driving[J]. (2025). arXiv preprint arXiv:2506.18084."},{"key":"24838_CR30","unstructured":"Meinert, N. & Lavin, A. Multivariate deep evidential regression[J]. arXiv preprint arXiv:2104.06135, (2021)."},{"key":"24838_CR31","doi-asserted-by":"crossref","unstructured":"Meinert, N., Gawlikowski, J. & Lavin, A. The unreasonable effectiveness of deep evidential regression[C]\/\/Proceedings of the AAAI Conference on Artificial Intelligence. 37(8): 9134\u20139142. (2023).","DOI":"10.1609\/aaai.v37i8.26096"},{"key":"24838_CR32","first-page":"18878","volume":"34","author":"B Liu","year":"2021","unstructured":"Liu, B. et al. Conflict-averse gradient descent for multi-task learning[J]. Adv. Neural. Inf. Process. Syst. 34, 18878\u201318890 (2021).","journal-title":"Adv. Neural. Inf. Process. Syst."},{"key":"24838_CR33","unstructured":"Mao, D. et al. Robust analysis of multi-task learning efficiency: new benchmarks on light-weighed backbones and effective measurement of multi-task learning challenges by feature disentanglement[J]. (2024). arXiv preprint arXiv:2402.03557."},{"key":"24838_CR34","doi-asserted-by":"crossref","unstructured":"Liu, X. et al. Anti-interference filter generator based on MGDA-UB strategy for V2X[C]\/\/2023 IEEE\/CIC international conference on communications in China (ICCC Workshops). IEEE, : 1\u20136. (2023).","DOI":"10.1109\/ICCCWorkshops57813.2023.10233797"},{"key":"24838_CR35","doi-asserted-by":"crossref","unstructured":"Qin, X., Wang, X. & Yan, J. Towards consistent multi-task learning: unlocking the potential of task-specific parameters[C]\/\/proceedings of the computer vision and pattern recognition conference. : 10067\u201310076. (2025).","DOI":"10.1109\/CVPR52734.2025.00941"},{"issue":"1","key":"24838_CR36","doi-asserted-by":"publisher","first-page":"3","DOI":"10.1365\/s13291-023-00275-x","volume":"126","author":"S Pokutta","year":"2024","unstructured":"Pokutta, S. The frank-wolfe algorithm: a short introduction[J]. Jahresbericht Der Deutschen Mathematiker-Vereinigung. 126 (1), 3\u201335 (2024).","journal-title":"Jahresbericht Der Deutschen Mathematiker-Vereinigung"},{"key":"24838_CR37","unstructured":"Nazykov, R. et al. Stochastic frank-wolfe: Unified analysis and zoo of special cases[C]\/\/International conference on artificial intelligence and statistics. PMLR, : 4870\u20134878. (2024)."}],"container-title":["Scientific Reports"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.nature.com\/articles\/s41598-025-24838-3.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-025-24838-3","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-025-24838-3.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,11,20]],"date-time":"2025-11-20T10:43:03Z","timestamp":1763635383000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.nature.com\/articles\/s41598-025-24838-3"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,11,20]]},"references-count":37,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2025,12]]}},"alternative-id":["24838"],"URL":"https:\/\/doi.org\/10.1038\/s41598-025-24838-3","relation":{},"ISSN":["2045-2322"],"issn-type":[{"value":"2045-2322","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,11,20]]},"assertion":[{"value":"26 August 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"15 October 2025","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"20 November 2025","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"The authors declare no competing interests.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}},{"value":"This article does not contain any studies with human participants or animals performed by any of the authors. All methods were performed in accordance with relevant guidelines and regulations.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Ethics statements"}}],"article-number":"40993"}}