{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,10,4]],"date-time":"2025-10-04T00:39:30Z","timestamp":1759538370290,"version":"build-2065373602"},"reference-count":31,"publisher":"Association for Computing Machinery (ACM)","issue":"5s","funder":[{"name":"National Research Foundation, Singapore under its AI Singapore Programme","award":["AISG4-GC-2023-006-1B"],"award-info":[{"award-number":["AISG4-GC-2023-006-1B"]}]},{"name":"International Conference on Embedded Software"}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Trans. Embed. Comput. Syst."],"published-print":{"date-parts":[[2025,11,30]]},"abstract":"<jats:p>\n            To maintain driving safety, the execution of neural network-based autonomous driving pipelines must meet the dynamic deadlines in response to the changing environment and vehicle\u2019s velocity. To this end, this article proposes a real-time neural architecture adaptation approach, called TimelyNet, which uses a\n            <jats:italic toggle=\"yes\">supernet<\/jats:italic>\n            to replace the most compute-intensive neural network module in an existing end-to-end autonomous driving pipeline. From the supernet, TimelyNet samples\n            <jats:italic toggle=\"yes\">subnets<\/jats:italic>\n            with varying inference latency levels to meet the dynamic deadlines during run-time driving without fine-tuning. Specifically, TimelyNet employs a one-shot prediction method that jointly uses a lookup table and an invertible neural network to periodically determine the optimal hyperparameters of a subnet to meet its execution deadline while achieving the highest possible accuracy. The lookup table stores multiple subnet architectures with different latencies, while the invertible neural network models the distribution of the optimal subnet architecture given the latency. Extensive evaluation based on hardware-in-the-loop CARLA simulations shows that TimelyNet-integrated driving pipelines achieve the best driving safety, characterized by the lowest wrong-lane driving rate and zero collisions, compared with several baselines, including the state-of-the-art driving pipelines.\n          <\/jats:p>","DOI":"10.1145\/3762652","type":"journal-article","created":{"date-parts":[[2025,8,26]],"date-time":"2025-08-26T11:26:21Z","timestamp":1756207581000},"page":"1-23","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["TimelyNet: Adaptive Neural Architecture for Autonomous Driving with Dynamic Deadline"],"prefix":"10.1145","volume":"24","author":[{"ORCID":"https:\/\/orcid.org\/0009-0003-8623-0172","authenticated-orcid":false,"given":"Jiale","family":"Chen","sequence":"first","affiliation":[{"name":"College of Computing and Data Science, Nanyang Technological University","place":["Singapore, Singapore"]}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0115-8726","authenticated-orcid":false,"given":"Duc Van","family":"Le","sequence":"additional","affiliation":[{"name":"College of Computing and Data Science, Nanyang Technological University","place":["Singapore, Singapore"]}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1591-2526","authenticated-orcid":false,"given":"Yuanchun","family":"Li","sequence":"additional","affiliation":[{"name":"Institute for AI Industry Research (AIR), Tsinghua University","place":["Beijing, China"]}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7352-8955","authenticated-orcid":false,"given":"Yunxin","family":"Liu","sequence":"additional","affiliation":[{"name":"Institute for AI Industry Research (AIR), Tsinghua University","place":["Beijing, China"]}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8441-9973","authenticated-orcid":false,"given":"Rui","family":"Tan","sequence":"additional","affiliation":[{"name":"College of Computing and Data Science, Nanyang Technological University","place":["Singapore, Singapore"]}]}],"member":"320","published-online":{"date-parts":[[2025,9,26]]},"reference":[{"key":"e_1_3_1_2_2","volume-title":"Framework for Easily Invertible Architectures (FrEIA)","author":"Ardizzone Lynton","year":"2018","unstructured":"Lynton Ardizzone, Till Bungert, Felix Draxler, Ullrich K\u00f6the, Jakob Kruse, Robert Schmier, and Peter Sorrenson. 2018-2022. Framework for Easily Invertible Architectures (FrEIA). Retrieved from https:\/\/github.com\/vislearn\/FrEIA"},{"key":"e_1_3_1_3_2","doi-asserted-by":"crossref","unstructured":"Lynton Ardizzone Jakob Kruse Carsten L\u00fcth Niels Bracher Carsten Rother and Ullrich K\u00f6the. 2020. Conditional invertible neural networks for diverse image-to-image translation. In DAGM German Conference on Pattern Recognition. Cham: Springer International Publishing 373\u2013387.","DOI":"10.1007\/978-3-030-71278-5_27"},{"key":"e_1_3_1_4_2","volume-title":"Proceedings of the International Conference on Learning Representations (ICLR)","author":"Cai Han","year":"2020","unstructured":"Han Cai, Chuang Gan, Tianzhe Wang, Zhekai Zhang, and Song Han. 2020. Once for all: Train one network and specialize it for efficient deployment. In Proceedings of the International Conference on Learning Representations (ICLR)."},{"key":"e_1_3_1_5_2","unstructured":"CARLA. 2023. CARLA Autonomous driving Leaderboard. Retrieved from https:\/\/leaderboard.carla.org\/"},{"key":"e_1_3_1_6_2","doi-asserted-by":"publisher","DOI":"10.1109\/TIV.2023.3318070"},{"key":"e_1_3_1_7_2","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2022.3200245"},{"key":"e_1_3_1_8_2","first-page":"5972","volume-title":"Proceedings of the IEEE\/CVF International Conference on Computer Vision (ICCV)","author":"Chu Xiangxiang","year":"2023","unstructured":"Xiangxiang Chu, Shun Lu, Xudong Li, and Bo Zhang. 2023. Mixpath: A unified approach for one-shot neural architecture search. In Proceedings of the IEEE\/CVF International Conference on Computer Vision (ICCV). 5972\u20135981."},{"key":"e_1_3_1_9_2","volume-title":"Proceedings of the 17th European Conference on Computer Systems (EuroSys)","author":"Gog Ionel","year":"2022","unstructured":"Ionel Gog, Sukrit Kalra, Peter Schafhalter, Joseph E Gonzalez, and Ion Stoica. 2022. D3: A dynamic deadline-driven approach for building autonomous vehicles. In Proceedings of the 17th European Conference on Computer Systems (EuroSys)."},{"key":"e_1_3_1_10_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICRA48506.2021.9561747"},{"key":"e_1_3_1_11_2","first-page":"544","volume-title":"Proceedings of the European Conference on Computer Vision (ECCV)","author":"Guo Zichao","year":"2020","unstructured":"Zichao Guo, Xiangyu Zhang, Haoyuan Mu, Wen Heng, Zechun Liu, Yichen Wei, and Jian Sun. 2020. Single path one-shot neural architecture search with uniform sampling. In Proceedings of the European Conference on Computer Vision (ECCV). Springer, 544\u2013560."},{"key":"e_1_3_1_12_2","first-page":"406","volume-title":"Proceedings of the 27th Annual International Conference on Mobile Computing and Networking (MobiCom)","author":"Han Rui","year":"2021","unstructured":"Rui Han, Qinglong Zhang, Chi Harold Liu, Guoren Wang, Jian Tang, and Lydia Y Chen. 2021. Legodnn: Block-grained scaling of deep neural networks for mobile vision. In Proceedings of the 27th Annual International Conference on Mobile Computing and Networking (MobiCom). 406\u2013419."},{"key":"e_1_3_1_13_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"key":"e_1_3_1_14_2","doi-asserted-by":"publisher","DOI":"10.1145\/3626098"},{"key":"e_1_3_1_15_2","doi-asserted-by":"publisher","DOI":"10.1109\/TMC.2024.3462437"},{"key":"e_1_3_1_16_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.01712"},{"key":"e_1_3_1_17_2","doi-asserted-by":"publisher","DOI":"10.1145\/3173162.3173191"},{"key":"e_1_3_1_18_2","doi-asserted-by":"publisher","DOI":"10.1145\/3477006"},{"key":"e_1_3_1_19_2","first-page":"1333","volume-title":"Proceedings of the IEEE\/CVF International Conference on Computer Vision (ICCV)","author":"Sakuma Yuiko","year":"2023","unstructured":"Yuiko Sakuma, Masato Ishii, and Takuya Narihira. 2023. DetOFA: Efficient training of once-for-all networks for object detection using path filter. In Proceedings of the IEEE\/CVF International Conference on Computer Vision (ICCV). 1333\u20131342."},{"key":"e_1_3_1_20_2","first-page":"726","volume-title":"Proceedings of the Conference on Robot Learning (CORL)","author":"Shao Hao","year":"2023","unstructured":"Hao Shao, Letian Wang, Ruobing Chen, Hongsheng Li, and Yu Liu. 2023. Safety-enhanced autonomous driving using interpretable sensor fusion transformer. In Proceedings of the Conference on Robot Learning (CORL). PMLR, 726\u2013737."},{"key":"e_1_3_1_21_2","doi-asserted-by":"publisher","DOI":"10.1145\/3569472"},{"key":"e_1_3_1_22_2","doi-asserted-by":"publisher","DOI":"10.1109\/MM.2020.2975764"},{"key":"e_1_3_1_23_2","unstructured":"Baidu Apollo team. 2017. Apollo: Open Source Autonomous Driving. Retrieved from https:\/\/github.com\/ApolloAuto\/apollo"},{"key":"e_1_3_1_24_2","first-page":"13029","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","author":"Wang Chien-Yao","year":"2021","unstructured":"Chien-Yao Wang, Alexey Bochkovskiy, and Hong-Yuan Mark Liao. 2021. Scaled-YOLOv4: Scaling cross stage partial network. In Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 13029\u201313038."},{"key":"e_1_3_1_25_2","doi-asserted-by":"crossref","first-page":"118","DOI":"10.1109\/RTSS59052.2023.00020","volume-title":"Proceedings of the 2023 IEEE Real-Time Systems Symposium (RTSS)","author":"Wang Ruiqi","year":"2023","unstructured":"Ruiqi Wang, Hanyang Liu, Jiaming Qiu, Moran Xu, Roch Gu\u00e9rin, and Chenyang Lu. 2023. Progressive neural compression for adaptive image offloading under timing constraints. In Proceedings of the 2023 IEEE Real-Time Systems Symposium (RTSS). IEEE, 118\u2013130."},{"key":"e_1_3_1_26_2","doi-asserted-by":"publisher","DOI":"10.1145\/3570361.3592529"},{"key":"e_1_3_1_27_2","volume-title":"Proceedings of the Neural Information Processing Systems (NeurIPS)","author":"Wu Penghao","year":"2022","unstructured":"Penghao Wu, Xiaosong Jia, Li Chen, Junchi Yan, Hongyang Li, and Yu Qiao. 2022. Trajectory-guided control prediction for end-to-end autonomous driving: A simple yet strong baseline. In Proceedings of the Neural Information Processing Systems (NeurIPS)."},{"key":"e_1_3_1_28_2","first-page":"126","volume-title":"Proceedings of the European Conference on Computer Vision (ECCV)","author":"Xiao Mingqing","year":"2020","unstructured":"Mingqing Xiao, Shuxin Zheng, Chang Liu, Yaolong Wang, Di He, Guolin Ke, Jiang Bian, Zhouchen Lin, and Tie-Yan Liu. 2020. Invertible image rescaling. In Proceedings of the European Conference on Computer Vision (ECCV). Springer, 126\u2013144."},{"key":"e_1_3_1_29_2","doi-asserted-by":"publisher","DOI":"10.1145\/3570503"},{"key":"e_1_3_1_30_2","first-page":"1999","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","author":"You Shan","year":"2020","unstructured":"Shan You, Tao Huang, Mingmin Yang, Fei Wang, Chen Qian, and Changshui Zhang. 2020. Greedynas: Towards fast one-shot nas with greedy supernet. In Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 1999\u20132008."},{"key":"e_1_3_1_31_2","volume-title":"Proceedings of the IEEE\/CVF International Conference on Computer Vision (ICCV)","author":"Zhang Zhejun","year":"2021","unstructured":"Zhejun Zhang, Alexander Liniger, Dengxin Dai, Fisher Yu, and Luc Van Gool. 2021. End-to-end urban driving by imitating a reinforcement learning coach. In Proceedings of the IEEE\/CVF International Conference on Computer Vision (ICCV)."},{"key":"e_1_3_1_32_2","first-page":"1","volume-title":"Proceedings of the 39th International Conference on Computer-Aided Design (ICCAD)","author":"Zhu Qi","year":"2020","unstructured":"Qi Zhu, Wenchao Li, Hyoseung Kim, Yecheng Xiang, Kacper Wardega, Zhilu Wang, Yixuan Wang, Hengyi Liang, Chao Huang, Jiameng Fan, and Hyunjong Choi. 2020. Know the unknowns: Addressing disturbances and uncertainties in autonomous systems. In Proceedings of the 39th International Conference on Computer-Aided Design (ICCAD). 1\u20139."}],"container-title":["ACM Transactions on Embedded Computing Systems"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3762652","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,3]],"date-time":"2025-10-03T14:06:08Z","timestamp":1759500368000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3762652"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,9,26]]},"references-count":31,"journal-issue":{"issue":"5s","published-print":{"date-parts":[[2025,11,30]]}},"alternative-id":["10.1145\/3762652"],"URL":"https:\/\/doi.org\/10.1145\/3762652","relation":{},"ISSN":["1539-9087","1558-3465"],"issn-type":[{"type":"print","value":"1539-9087"},{"type":"electronic","value":"1558-3465"}],"subject":[],"published":{"date-parts":[[2025,9,26]]},"assertion":[{"value":"2025-08-11","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2025-08-12","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2025-09-26","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}