TEM^3-Learning: Time-Efficient Multimodal Multi-Task Learning for Advanced Assistive Driving

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Liu, Wenzhuo, Qiao, Yicheng, Wang, Zhen, Guo, Qiannan, Chen, Zilong, Zhou, Meihua, Li, Xinran, Wang, Letian, Li, Zhiwei, Liu, Huaping, Wang, Wenshuo
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866908416505544704
author Liu, Wenzhuo
Qiao, Yicheng
Wang, Zhen
Guo, Qiannan
Chen, Zilong
Zhou, Meihua
Li, Xinran
Wang, Letian
Li, Zhiwei
Liu, Huaping
Wang, Wenshuo
author_facet Liu, Wenzhuo
Qiao, Yicheng
Wang, Zhen
Guo, Qiannan
Chen, Zilong
Zhou, Meihua
Li, Xinran
Wang, Letian
Li, Zhiwei
Liu, Huaping
Wang, Wenshuo
contents Multi-task learning (MTL) can advance assistive driving by exploring inter-task correlations through shared representations. However, existing methods face two critical limitations: single-modality constraints limiting comprehensive scene understanding and inefficient architectures impeding real-time deployment. This paper proposes TEM^3-Learning (Time-Efficient Multimodal Multi-task Learning), a novel framework that jointly optimizes driver emotion recognition, driver behavior recognition, traffic context recognition, and vehicle behavior recognition through a two-stage architecture. The first component, the mamba-based multi-view temporal-spatial feature extraction subnetwork (MTS-Mamba), introduces a forward-backward temporal scanning mechanism and global-local spatial attention to efficiently extract low-cost temporal-spatial features from multi-view sequential images. The second component, the MTL-based gated multimodal feature integrator (MGMI), employs task-specific multi-gating modules to adaptively highlight the most relevant modality features for each task, effectively alleviating the negative transfer problem in MTL. Evaluation on the AIDE dataset, our proposed model achieves state-of-the-art accuracy across all four tasks, maintaining a lightweight architecture with fewer than 6 million parameters and delivering an impressive 142.32 FPS inference speed. Rigorous ablation studies further validate the effectiveness of the proposed framework and the independent contributions of each module. The code is available on https://github.com/Wenzhuo-Liu/TEM3-Learning.
format Preprint
id arxiv_https___arxiv_org_abs_2506_18084
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle TEM^3-Learning: Time-Efficient Multimodal Multi-Task Learning for Advanced Assistive Driving
Liu, Wenzhuo
Qiao, Yicheng
Wang, Zhen
Guo, Qiannan
Chen, Zilong
Zhou, Meihua
Li, Xinran
Wang, Letian
Li, Zhiwei
Liu, Huaping
Wang, Wenshuo
Computer Vision and Pattern Recognition
Multi-task learning (MTL) can advance assistive driving by exploring inter-task correlations through shared representations. However, existing methods face two critical limitations: single-modality constraints limiting comprehensive scene understanding and inefficient architectures impeding real-time deployment. This paper proposes TEM^3-Learning (Time-Efficient Multimodal Multi-task Learning), a novel framework that jointly optimizes driver emotion recognition, driver behavior recognition, traffic context recognition, and vehicle behavior recognition through a two-stage architecture. The first component, the mamba-based multi-view temporal-spatial feature extraction subnetwork (MTS-Mamba), introduces a forward-backward temporal scanning mechanism and global-local spatial attention to efficiently extract low-cost temporal-spatial features from multi-view sequential images. The second component, the MTL-based gated multimodal feature integrator (MGMI), employs task-specific multi-gating modules to adaptively highlight the most relevant modality features for each task, effectively alleviating the negative transfer problem in MTL. Evaluation on the AIDE dataset, our proposed model achieves state-of-the-art accuracy across all four tasks, maintaining a lightweight architecture with fewer than 6 million parameters and delivering an impressive 142.32 FPS inference speed. Rigorous ablation studies further validate the effectiveness of the proposed framework and the independent contributions of each module. The code is available on https://github.com/Wenzhuo-Liu/TEM3-Learning.
title TEM^3-Learning: Time-Efficient Multimodal Multi-Task Learning for Advanced Assistive Driving
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2506.18084