Context-Aware Multi-Task Learning for Traffic Scene Recognition in Autonomous Vehicles

被引:0
|
作者
Lee, Younkwan [1 ]
Jeon, Jihyo [1 ]
Yu, Jongmin [1 ,2 ]
Jeon, Moongu [1 ]
机构
[1] Gwangju Inst Sci & Technol GIST, Machine Learning & Vis Lab, Gwangju 61005, South Korea
[2] Curtin Univ, Bentley, WA 6102, Australia
基金
新加坡国家研究基金会;
关键词
CLASSIFICATION; TIME; FEATURES; SPACE;
D O I
10.1109/iv47402.2020.9304708
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Traffic scene recognition, which requires various visual classification tasks, is a critical ingredient in autonomous vehicles. However, most existing approaches treat each relevant task independently from one another, never considering the entire system as a whole. Because of this, they are limited to utilizing a task-specific set of features for all possible tasks of inference-time, which ignores the capability to leverage common task-invariant contextual knowledge for the task at hand. To address this problem, we propose an algorithm to jointly learn the task-specific and shared representations by adopting a multi-task learning network. Specifically, we present a lower bound for the mutual information constraint between shared feature embedding and input that is considered to be able to extract common contextual information across tasks while preserving essential information of each task jointly. The learned representations capture richer contextual information without additional task-specific network. Extensive experiments on the large-scale dataset HSD demonstrate the effectiveness and superiority of our network over state-of-the-art methods.
引用
收藏
页码:723 / 730
页数:8
相关论文
共 50 条
  • [41] Scalable Multi-Task Imitation Learning with Autonomous Improvement
    Singh, Avi
    Jang, Eric
    Irpan, Alexander
    Kappler, Daniel
    Dalal, Murtaza
    Levinev, Sergey
    Khansari, Mohi
    Finn, Chelsea
    [J]. 2020 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA), 2020, : 2167 - 2173
  • [42] Optimal Configuration of Multi-Task Learning for Autonomous Driving
    Jun, Woomin
    Son, Minjun
    Yoo, Jisang
    Lee, Sungjin
    [J]. SENSORS, 2023, 23 (24)
  • [43] Scale-Aware Task Message Transferring for Multi-Task Learning
    Sirejiding, Shalayiding
    Lu, Yuxiang
    Lu, Hongtao
    Ding, Yue
    [J]. 2023 IEEE INTERNATIONAL CONFERENCE ON MULTIMEDIA AND EXPO, ICME, 2023, : 1859 - 1864
  • [44] Multitask Learning of Compact Semantic Codebooks for Context-Aware Scene Modeling
    Wang, Botao
    Xiong, Hongkai
    Lin, Weiyao
    Zou, Junni
    Zheng, Yuan F.
    [J]. IEEE TRANSACTIONS ON IMAGE PROCESSING, 2016, 25 (11) : 5411 - 5426
  • [45] E-HoA: A Distributed Layered Architecture for Context-aware Autonomous Vehicles
    Ilie, Jean-Michel
    Chaouche, Ahmed-Chawki
    Pecheux, Francois
    [J]. 11TH INTERNATIONAL CONFERENCE ON AMBIENT SYSTEMS, NETWORKS AND TECHNOLOGIES (ANT) / THE 3RD INTERNATIONAL CONFERENCE ON EMERGING DATA AND INDUSTRY 4.0 (EDI40) / AFFILIATED WORKSHOPS, 2020, 170 : 530 - 538
  • [46] Online Learning for Context-Aware Multi-User Package Delivery System with Unmanned Vehicles
    Zhang, Chonghao
    Zhou, Pan
    Liu, Guanghui
    Gong, Shimin
    Wang, Wei
    Wu, Dapeng
    [J]. ICC 2019 - 2019 IEEE INTERNATIONAL CONFERENCE ON COMMUNICATIONS (ICC), 2019,
  • [47] Task-Aware Dynamic Model Optimization for Multi-Task Learning
    Choi, Sujin
    Jin, Hyundong
    Kim, Eunwoo
    [J]. IEEE ACCESS, 2023, 11 : 137709 - 137717
  • [48] Stratified Multi-Task Learning for Robust Spotting of Scene Texts
    Dasgupta, Kinjal
    Das, Sudip
    Bhattacharya, Ujjwal
    [J]. 2020 25TH INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION (ICPR), 2021, : 3130 - 3137
  • [49] CATMI: Context-Aware Traffic Management at Autonomous Un-Signalized Intersections
    Vahdat-Nejad, Hamed
    Mohammadi, Tahereh
    Mansoor, Wathiq
    [J]. INTERNATIONAL JOURNAL OF INTELLIGENT TRANSPORTATION SYSTEMS RESEARCH, 2023, 21 (01) : 145 - 158
  • [50] CATMI: Context-Aware Traffic Management at Autonomous Un-Signalized Intersections
    Hamed Vahdat-Nejad
    Tahereh Mohammadi
    Wathiq Mansoor
    [J]. International Journal of Intelligent Transportation Systems Research, 2023, 21 : 145 - 158