BOUNDING BOX DISPARITY: 3D METRICS FOR OBJECT DETECTION WITH FULL DEGREE OF FREEDOM

被引：3
|
作者：
Adam, Michael G. [1 ]
Piccolrovazzi, Martin [1 ]
Eger, Sebastian [1 ]
Steinbach, Eckehard [1 ]
机构：
[1] Tech Univ Munich, Chair Media Technol & Munich Inst Robot & Machine, Munich, Germany
来源：
2022 IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING, ICIP | 2022年
关键词：
metric; object detection; 3D bounding box; intersection over union; volume-to-volume distance;
D O I：
10.1109/ICIP46576.2022.9897588
中图分类号：
TP18 [人工智能理论];
学科分类号：
081104 ; 0812 ; 0835 ; 1405 ;
摘要：
The most popular evaluation metric for object detection in 2D images is Intersection over Union (IoU). Existing implementations of the IoU metric for 3D object detection usually neglect one or more degrees of freedom. In this paper, we first derive the analytic solution for three dimensional bounding boxes. As a second contribution, a closed-form solution of the volume-to-volume distance is derived. Finally, the Bounding Box Disparity is proposed as a combined positive continuous metric. We provide open source implementations of the three metrics as standalone python functions, as well as extensions to the Open3D library and as ROS nodes.
引用
收藏
页码：1491 / 1495
页数：5
相关论文
共 50 条
  • [1] Monocular 3D Object Detection with Bounding Box Denoising in 3D by Perceiver
    Liu, Xianpeng
    Zheng, Ce
    Cheng, Kelvin
    Xue, Nan
    Qi, Guo-Jun
    Wu, Tianfu
    2023 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION, ICCV, 2023, : 6413 - 6423
  • [2] 3D object Classification using Bounding box
    Malwe, Gauri
    Kshirsagar, Deepak
    Madkaikar, Ashish
    2018 FOURTH INTERNATIONAL CONFERENCE ON COMPUTING COMMUNICATION CONTROL AND AUTOMATION (ICCUBEA), 2018,
  • [3] OBMO: One Bounding Box Multiple Objects for Monocular 3D Object Detection
    Huang, Chenxi
    He, Tong
    Ren, Haidong
    Wang, Wenxiao
    Lin, Binbin
    Cai, Deng
    IEEE TRANSACTIONS ON IMAGE PROCESSING, 2023, 32 : 6570 - 6581
  • [4] A review on object pose recovery: From 3D bounding box detectors to full 6D pose estimators
    Sahin, Caner
    Garcia-Hernando, Guillermo
    Sock, Juil
    Kim, Tae-Kyun
    IMAGE AND VISION COMPUTING, 2020, 96
  • [5] Estimation of a 3D Bounding Box for a Segmented Object Region in a Single Image
    Jung, Sunghoon
    Kim, Minhwan
    IEICE TRANSACTIONS ON INFORMATION AND SYSTEMS, 2014, E97D (11): : 2919 - 2934
  • [6] A 3D Multi-Object Tracking Based on Bounding Box and Depth
    Feng, Chun-Hao
    Chang, Jen-Yuan
    PROCEEDINGS OF THE ASME 2021 30TH CONFERENCE ON INFORMATION STORAGE AND PROCESSING SYSTEMS (ISPS2021), 2021,
  • [7] 3D Vehicle Object Tracking Algorithm Based on Bounding Box Similarity Measurement
    Cheng, Xin
    Zhou, Jingmei
    Liu, Peiyuan
    Zhao, Xiangmo
    Wang, Hongfei
    IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, 2023, 24 (12) : 15844 - 15854
  • [8] NOVEL ANNOTATION AND METRICS FOR MANGROVE SPECIES CLASSIFICATION USING BOUNDING BOX OBJECT DETECTION
    Lim, Han Shen
    Lee, Yunli
    Eu, Kok Seng
    Lin, Mei-Hua
    Yap, Kian Meng
    Chia, Wai Chong
    2023 IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING, ICIP, 2023, : 2030 - 2034
  • [9] Representation Disparity-aware Distillation for 3D Object Detection
    Li, Yanjing
    Xu, Sheng
    Lin, Mingbao
    Yin, Jihao
    Zhang, Baochang
    Cao, Xianbin
    2023 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION, ICCV, 2023, : 6692 - 6701
  • [10] Vehicle Detection Using Point Cloud and 3D LIDAR Sensor to Draw 3D Bounding Box
    Gagana, H. S.
    Sunitha, N. R.
    Nishanth, K. N.
    COMPUTATIONAL VISION AND BIO-INSPIRED COMPUTING, 2020, 1108 : 983 - 992
12345