{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,5]],"date-time":"2026-08-05T08:49:13Z","timestamp":1785919753823,"version":"3.56.0"},"reference-count":47,"publisher":"MDPI AG","issue":"15","license":[{"start":{"date-parts":[[2026,8,3]],"date-time":"2026-08-03T00:00:00Z","timestamp":1785715200000},"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 presents the latent model-joint embedding predictive architecture (LM-JEPA), a resource-efficient collaborative perception framework for connected and autonomous vehicles that integrates latent predictive representation learning with lightweight multi-modal reasoning. Autonomous driving in urban and highway environments requires accurate scene understanding under strict latency, energy, and communication constraints, limiting the practicality of large language model (LLM) and vision\u2013language model (VLM)-based approaches in edge deployments. To address this, LM-JEPA encodes heterogeneous inputs including camera, LiDAR, radar, and map data into a unified latent space using a joint embedding predictive architecture, enabling efficient perception and reasoning without token-level inference. Unlike existing latent-space learning approaches that primarily learn predictive visual embeddings for single-modal perception, the proposed framework integrates multi-modal latent reasoning and adaptive sensor fusion to support collaborative perception under resource-constrained vehicular edge environments. The collaborative perception framework introduces a context-adaptive multi-modal fusion mechanism that dynamically weights sensor and model contributions, along with selective latent transmission and adaptive decoding for resource-aware operation. A lightweight VLM is integrated with an edge-assisted vehicular pipeline to support real-time on-vehicle inference with adaptive offloading based on latency and energy constraints, while a latent-space reasoning module enables cooperative decision-making. Experiments on BDD100K and nuScenes-QA show that LM-JEPA improves perception accuracy by 5% and reduces latency by approximately 7% over LLM and VLM baselines, while achieving up to 25% improvement in scene understanding, 20% higher intersection success rates, improved highway merging, and approximately 15% reduction in the transmitted model parameters.<\/jats:p>","DOI":"10.3390\/s26154894","type":"journal-article","created":{"date-parts":[[2026,8,3]],"date-time":"2026-08-03T11:50:21Z","timestamp":1785757821000},"page":"4894","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Object Detection and Scene Perception for Connected and Autonomous Vehicles Using LM-JEPA"],"prefix":"10.3390","volume":"26","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-8728-5146","authenticated-orcid":false,"given":"Abhishek","family":"Gupta","sequence":"first","affiliation":[{"name":"Faculty of Computer Science and Technology, Algoma University, Brampton, ON L6V 1A3, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8591-026X","authenticated-orcid":false,"given":"Ajmery","family":"Sultana","sequence":"additional","affiliation":[{"name":"Faculty of Computer Science and Technology, Algoma University, Brampton, ON L6V 1A3, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2026,8,3]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"397","DOI":"10.1109\/OJVT.2024.3369691","article-title":"Advanced Learning Technologies for Intelligent Transportation Systems: Prospects and Challenges","volume":"5","author":"Khalil","year":"2024","journal-title":"IEEE Open J. Veh. Technol."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"762","DOI":"10.1109\/TNSE.2025.3587238","article-title":"Foundation Model Empowered Synesthesia of Machines (SoM): AI-native Intelligent Multi-Modal Sensing-Communication Integration","volume":"13","author":"Cheng","year":"2025","journal-title":"IEEE Trans. Netw. Sci. Eng."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"1955","DOI":"10.1109\/COMST.2024.3465447","article-title":"Large Language Model (LLM) for Telecommunications: A Comprehensive Survey on Principles, Key Techniques, and Opportunities","volume":"27","author":"Zhou","year":"2025","journal-title":"IEEE Commun. Surv. Tutor."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"98","DOI":"10.1109\/MWC.001.2400384","article-title":"Large Language Models for Wireless Networks: An Overview from the Prompt Engineering Perspective","volume":"32","author":"Zhou","year":"2025","journal-title":"IEEE Wirel. Commun."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"615","DOI":"10.1109\/OJITS.2026.3665677","article-title":"LLM and AI Agents for Autonomous Systems: A Survey of Applications, Datasets, and Security Challenges","volume":"7","author":"Ferrag","year":"2026","journal-title":"IEEE Open J. Intell. Transp. Syst."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"187","DOI":"10.1109\/TITS.2025.3628969","article-title":"Large (Vision) Language Models for Autonomous Vehicles: Current Trends and Future Directions","volume":"27","author":"Tian","year":"2026","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"55344","DOI":"10.1109\/JIOT.2025.3624038","article-title":"Multi-Agent Collaborative Decision-Making Using Small Vision-Language Models for Autonomous Driving","volume":"12","author":"Chi","year":"2025","journal-title":"IEEE Internet Things J."},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"108076","DOI":"10.1109\/ACCESS.2025.3580641","article-title":"Efficacy of Autonomous Vehicle\u2019s Adaptive Decision-Making Based on Large Language Models Across Multiple Driving Scenarios","volume":"13","author":"Xiong","year":"2025","journal-title":"IEEE Access"},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"32701","DOI":"10.1109\/JIOT.2025.3579032","article-title":"Vision-Language Models for Edge Networks: A Comprehensive Survey","volume":"12","author":"Sharshar","year":"2025","journal-title":"IEEE Internet Things J."},{"key":"ref_10","doi-asserted-by":"crossref","unstructured":"Wang, J., Ren, H., Zhu, X., and Ma, Z. (2025). Enhancing Autonomous Vehicle Decision-Making Through Policy Transfer With Large Language Model. IEEE Trans. Intell. Transp. Syst., 1\u201310.","DOI":"10.1109\/TITS.2025.3565762"},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"5531","DOI":"10.1109\/TVT.2025.3623328","article-title":"Game-Theoretic Decision-Making for Autonomous Vehicles at Unsignalized Intersections under Communication Interferences: A Novel Risk-Adaptive Approach","volume":"75","author":"Zhu","year":"2025","journal-title":"IEEE Trans. Veh. Technol."},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"6601","DOI":"10.1109\/TITS.2025.3534836","article-title":"Enhancing the Collaborative Decision-Making Performance of Connected and Autonomous Vehicles: A Multi-Modal Failure-Aware Graph Representation Approach","volume":"26","author":"Liu","year":"2025","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"1450","DOI":"10.1109\/TIV.2023.3327715","article-title":"DriveLLM: Charting the Path Toward Full Autonomous Driving With Large Language Models","volume":"9","author":"Cui","year":"2024","journal-title":"IEEE Trans. Intell. Veh."},{"key":"ref_14","doi-asserted-by":"crossref","unstructured":"Zheng, Y., Xing, Z., Zhang, Q., Jin, B., Li, P., Zheng, Y., Xia, Z., Chen, Y., and Zhao, D. (2026). PlanAgent: A Multi-modal Large Language Agent for Closed-loop Vehicle Motion Planning. IEEE Trans. Cogn. Dev. Syst., 1\u201314.","DOI":"10.1109\/TCDS.2026.3664120"},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"1950","DOI":"10.1109\/TSE.2025.3569086","article-title":"TARGET: Traffic Rule-Based Test Generation for Autonomous Driving via Validated LLM-Guided Knowledge Extraction","volume":"51","author":"Deng","year":"2025","journal-title":"IEEE Trans. Softw. Eng."},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"16630","DOI":"10.1109\/TVT.2025.3572440","article-title":"Adaptive Resource Allocation Optimization Using Large Language Models in Dynamic Wireless Environments","volume":"74","author":"Noh","year":"2025","journal-title":"IEEE Trans. Veh. Technol."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"5799","DOI":"10.1109\/OJCOMS.2024.3456549","article-title":"LLM-Based Edge Intelligence: A Comprehensive Survey on Architectures, Applications, Security and Trustworthiness","volume":"5","author":"Friha","year":"2024","journal-title":"IEEE Open J. Commun. Soc."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"1470","DOI":"10.1109\/TITS.2024.3507536","article-title":"Steering Angle-Guided Multimodal Fusion Lane Detection for Autonomous Driving","volume":"26","author":"Gong","year":"2025","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"19833","DOI":"10.1109\/TITS.2025.3612873","article-title":"Fuse It or Lose It? Analyzing the Effects of Sensor Diversity on Multimodal Ensembles for Autonomous Vehicle Perception","volume":"26","author":"Mortlock","year":"2025","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"90144","DOI":"10.1109\/ACCESS.2025.3570711","article-title":"Camscribe: Enhanced Dashcam Video Descriptions Through Multimodal Spatiotemporal and Object Detection for Autonomous Vehicles","volume":"13","author":"Rafiq","year":"2025","journal-title":"IEEE Access"},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"3642","DOI":"10.1109\/TVT.2025.3608811","article-title":"DriveSOTIF: Advancing SOTIF Through Multimodal Large Language Models","volume":"75","author":"Huang","year":"2025","journal-title":"IEEE Trans. Veh. Technol."},{"key":"ref_22","first-page":"1","article-title":"Unseen From Seen: Rewriting Observation-Instruction Using Foundation Models for Augmenting Vision-Language Navigation","volume":"1","author":"Wei","year":"2025","journal-title":"IEEE Trans. Neural Netw. Learn. Syst."},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"44382","DOI":"10.1038\/s41598-025-28120-4","article-title":"Graph neural network\u2013driven text classification for fire-door defect inspection in pre-completion construction","volume":"15","author":"Wang","year":"2025","journal-title":"Sci. Rep."},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"43910","DOI":"10.1038\/s41598-025-27648-9","article-title":"Development of an automated transformer-based text analysis framework for monitoring fire door defects in buildings","volume":"15","author":"Wang","year":"2025","journal-title":"Sci. Rep."},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"132547","DOI":"10.1109\/ACCESS.2025.3589319","article-title":"Large Language Models in Transportation: A Comprehensive Bibliometric Analysis of Emerging Trends, Challenges, and Future Research","volume":"13","author":"Hassan","year":"2025","journal-title":"IEEE Access"},{"key":"ref_26","doi-asserted-by":"crossref","first-page":"95851","DOI":"10.1109\/ACCESS.2025.3573955","article-title":"A Comprehensive Overview and Analysis of Large Language Models: Trends and Challenges","volume":"13","author":"Mohammed","year":"2025","journal-title":"IEEE Access"},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"22","DOI":"10.1109\/TAI.2025.3569516","article-title":"Inadequacies of Large Language Model Benchmarks in the Era of Generative Artificial Intelligence","volume":"7","author":"McIntosh","year":"2026","journal-title":"IEEE Trans. Artif. Intell."},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"3815","DOI":"10.1109\/TCCN.2025.3620286","article-title":"A Weighted Byzantine Fault Tolerance Consensus Driven Trusted Multiple Large Language Models Network","volume":"12","author":"Luo","year":"2026","journal-title":"IEEE Trans. Cogn. Commun. Netw."},{"key":"ref_29","doi-asserted-by":"crossref","first-page":"11385","DOI":"10.1109\/TCOMM.2025.3592598","article-title":"On the Convergence of Large Language Model Optimizer for Black-Box Network Management","volume":"73","author":"Lee","year":"2025","journal-title":"IEEE Trans. Commun."},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"835","DOI":"10.1109\/TMLCN.2025.3585845","article-title":"LLM4WM: Adapting LLM for Wireless Multi-Tasking","volume":"3","author":"Liu","year":"2025","journal-title":"IEEE Trans. Mach. Learn. Commun. Netw."},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"167660","DOI":"10.1109\/ACCESS.2025.3612989","article-title":"Large Language Model-Based Functional Scenario Generation for Automated Vehicle Safety Evaluation Using Vehicle and Pedestrian Traffic Accident Data","volume":"13","author":"Kang","year":"2025","journal-title":"IEEE Access"},{"key":"ref_32","doi-asserted-by":"crossref","first-page":"770","DOI":"10.1109\/TITS.2024.3486324","article-title":"SCNet3D: Rethinking the Feature Extraction Process of Pillar-Based 3D Object Detection","volume":"26","author":"Li","year":"2025","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"ref_33","doi-asserted-by":"crossref","first-page":"3208","DOI":"10.1109\/TSE.2024.3478317","article-title":"Automated Commit Message Generation with Large Language Models: An Empirical Study and Beyond","volume":"50","author":"Xue","year":"2024","journal-title":"IEEE Trans. Softw. Eng."},{"key":"ref_34","doi-asserted-by":"crossref","first-page":"1191","DOI":"10.1109\/TNSE.2025.3592367","article-title":"Can Large Language Models Be Trusted as Evolutionary Optimizers for Network-Structured Combinatorial Problems?","volume":"13","author":"Zhao","year":"2026","journal-title":"IEEE Trans. Netw. Sci. Eng."},{"key":"ref_35","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1109\/TMC.2025.3570376","article-title":"Enhancing Stability and Resource Efficiency in LLM Training for Edge-Assisted Mobile Systems","volume":"25","author":"Liu","year":"2026","journal-title":"IEEE Trans. Mob. Comput."},{"key":"ref_36","doi-asserted-by":"crossref","first-page":"6332","DOI":"10.1109\/TPAMI.2026.3656175","article-title":"Evaluating and Mitigating Relationship Hallucinations in Large Vision-Language Models","volume":"48","author":"Wu","year":"2026","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"ref_37","doi-asserted-by":"crossref","first-page":"3242","DOI":"10.1109\/TMM.2026.3651099","article-title":"Mitigating Hallucinations in Large Vision-Language Models via Visual-Enhanced Contrastive Decoding","volume":"28","author":"Qiang","year":"2026","journal-title":"IEEE Trans. Multimed."},{"key":"ref_38","doi-asserted-by":"crossref","first-page":"7380","DOI":"10.1109\/TMM.2025.3599076","article-title":"Mitigating Hallucinations in Large Vision-Language Models via Reasoning Uncertainty-Guided Refinement","volume":"27","author":"Li","year":"2025","journal-title":"IEEE Trans. Multimed."},{"key":"ref_39","doi-asserted-by":"crossref","first-page":"358","DOI":"10.1109\/TITS.2025.3625700","article-title":"Hallucination Elimination and Text Annotation Framework for Large Vision-Language Models in Traffic Scenarios","volume":"27","author":"Fan","year":"2026","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"ref_40","doi-asserted-by":"crossref","first-page":"2945","DOI":"10.1109\/LSP.2024.3480043","article-title":"Towards Hybrid Quantum-Classical Deep Learning Architecture for Indoor-Outdoor Detection Using QCNN-LSTM and Cluster State Signal Processing","volume":"31","author":"Dastagir","year":"2024","journal-title":"IEEE Signal Process. Lett."},{"key":"ref_41","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1109\/TSE.2023.3334955","article-title":"An Empirical Evaluation of Using Large Language Models for Automated Unit Test Generation","volume":"50","author":"Schafer","year":"2024","journal-title":"IEEE Trans. Softw. Eng."},{"key":"ref_42","doi-asserted-by":"crossref","first-page":"933","DOI":"10.1109\/TG.2025.3564042","article-title":"Harnessing Language for Coordination: A Framework and Benchmark for LLM-Driven Multiagent Control","volume":"17","author":"Anne","year":"2025","journal-title":"IEEE Trans. Games"},{"key":"ref_43","doi-asserted-by":"crossref","first-page":"100057","DOI":"10.1016\/j.array.2021.100057","article-title":"Deep learning for object detection and scene perception in self-driving cars: Survey, challenges, and open issues","volume":"10","author":"Gupta","year":"2021","journal-title":"Array"},{"key":"ref_44","doi-asserted-by":"crossref","unstructured":"Padmasiri, H., Madurawe, R., Abeysinghe, C., and Meedeniya, D. (2020). Automated Vehicle Parking Occupancy Detection in Real-Time. Proceedings of the 2020 Moratuwa Engineering Research Conference (MERCon), Moratuwa, Sri Lanka, 28\u201330 July 2020, IEEE.","DOI":"10.1109\/MERCon50084.2020.9185199"},{"key":"ref_45","first-page":"4542","article-title":"NuScenes-QA: A Multi-Modal Visual Question Answering Benchmark for Autonomous Driving Scenario","volume":"38","author":"Qian","year":"2024","journal-title":"Proc. Conf. Artif. Intell."},{"key":"ref_46","doi-asserted-by":"crossref","unstructured":"Yu, F., Chen, H., Wang, X., Xian, W., Chen, Y., Liu, F., Madhavan, V., and Darrell, T. (2020). BDD100K: A Diverse Driving Dataset for Heterogeneous Multitask Learning. Proceedings of the (IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Online), IEEE.","DOI":"10.1109\/CVPR42600.2020.00271"},{"key":"ref_47","doi-asserted-by":"crossref","first-page":"16609","DOI":"10.1109\/JIOT.2026.3660030","article-title":"V2X-JEPA: Self-Supervised Multi-Agent Joint Embedding Predictive Architecture for Robust Vehicle-to-Everything Perception","volume":"13","author":"Mayumu","year":"2026","journal-title":"IEEE Internet Things J."}],"container-title":["Sensors"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1424-8220\/26\/15\/4894\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,5]],"date-time":"2026-08-05T08:16:42Z","timestamp":1785917802000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1424-8220\/26\/15\/4894"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,3]]},"references-count":47,"journal-issue":{"issue":"15","published-online":{"date-parts":[[2026,8]]}},"alternative-id":["s26154894"],"URL":"https:\/\/doi.org\/10.3390\/s26154894","relation":{},"ISSN":["1424-8220"],"issn-type":[{"value":"1424-8220","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,3]]}}}