MonoATT: Online Monocular 3D Object Detection with Adaptive Token Transformer

被引:11
|
作者
Zhou, Yunsong [1 ]
Zhu, Hongzi [1 ]
Liu, Quan [1 ]
Chang, Shan [2 ]
Guo, Minyi [1 ]
机构
[1] Shanghai Jiao Tong Univ, Shanghai, Peoples R China
[2] Donghua Univ, Shanghai, Peoples R China
基金
上海市自然科学基金; 中国国家自然科学基金;
关键词
D O I
10.1109/CVPR52729.2023.01678
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Mobile monocular 3D object detection (Mono3D) (e.g., on a vehicle, a drone, or a robot) is an important yet challenging task. Existing transformer-based offline Mono3D models adopt grid-based vision tokens, which is suboptimal when using coarse tokens due to the limited available computational power. In this paper, we propose an online Mono3D framework, called MonoATT, which leverages a novel vision transformer with heterogeneous tokens of varying shapes and sizes to facilitate mobile Mono3D. The core idea of MonoATT is to adaptively assign finer tokens to areas of more significance before utilizing a transformer to enhance Mono3D. To this end, we first use prior knowledge to design a scoring network for selecting the most important areas of the image, and then propose a token clustering and merging network with an attention mechanism to gradually merge tokens around the selected areas in multiple stages. Finally, a pixel-level feature map is reconstructed from heterogeneous tokens before employing a SOTA Mono3D detector as the underlying detection core. Experiment results on the real-world KITTI dataset demonstrate that MonoATT can effectively improve the Mono3D accuracy for both near and far objects and guarantee low latency. MonoATT yields the best performance compared with the state-of-the-art methods by a large margin and is ranked number one on the KITTI 3D benchmark.
引用
收藏
页码:17493 / 17503
页数:11
相关论文
共 50 条
  • [31] A New Monocular 3D Object Detection with Neural Network
    Hong, Weijie
    Liu, Yiguang
    Zheng, Yunan
    Wang, Ying
    Shi, Xuelei
    PATTERN RECOGNITION AND COMPUTER VISION (PRCV 2018), PT IV, 2018, 11259 : 174 - 185
  • [32] 3D Visual Object Detection from Monocular Images
    Wang, Qiaosong
    Rasmussen, Christopher
    ADVANCES IN VISUAL COMPUTING, ISVC 2019, PT I, 2020, 11844 : 168 - 180
  • [33] Competition for roadside camera monocular 3D object detection
    Jia, Jinrang
    Shi, Yifeng
    Qu, Yuli
    Wang, Rui
    Xu, Xing
    Zhang, Hai
    NATIONAL SCIENCE REVIEW, 2023, 10 (06)
  • [34] Monocular 3D Object Detection with Depth from Motion
    Wang, Tai
    Pang, Jiangmiao
    Lin, Dahua
    COMPUTER VISION, ECCV 2022, PT IX, 2022, 13669 : 386 - 403
  • [35] Objects are Different: Flexible Monocular 3D Object Detection
    Zhang, Yunpeng
    Lu, Jiwen
    Zhou, Jie
    2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2021, 2021, : 3288 - 3297
  • [36] Monocular 3D object detection for construction scene analysis
    Shen, Jie
    Jiao, Lang
    Zhang, Cong
    Peng, Keran
    COMPUTER-AIDED CIVIL AND INFRASTRUCTURE ENGINEERING, 2024, 39 (09) : 1370 - 1389
  • [37] Delving into Localization Errors for Monocular 3D Object Detection
    Ma, Xinzhu
    Zhang, Yinmin
    Xu, Dan
    Zhou, Dongzhan
    Yi, Shuai
    Li, Haojie
    Ouyang, Wanli
    2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2021, 2021, : 4719 - 4728
  • [38] Shape-Aware Monocular 3D Object Detection
    Chen, Wei
    Zhao, Jie
    Zhao, Wan-Lei
    Wu, Song-Yuan
    IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, 2023, 24 (06) : 6416 - 6424
  • [39] Competition for roadside camera monocular 3D object detection
    Jinrang Jia
    Yifeng Shi
    Yuli Qu
    Rui Wang
    Xing Xu
    Hai Zhang
    NationalScienceReview, 2023, 10 (06) : 34 - 37
  • [40] MonoGRNet: A General Framework for Monocular 3D Object Detection
    Qin, Zengyi
    Wang, Jinglu
    Lu, Yan
    IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2022, 44 (09) : 5170 - 5184