Panoptic 3D Scene Reconstruction From a Single RGB Image

被引:0
|
作者
Dahnert, Manuel [1 ]
Hou, Ji [1 ]
Niessner, Matthias [1 ]
Dai, Angela [1 ]
机构
[1] Tech Univ Munich, Munich, Germany
基金
欧洲研究理事会;
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Understanding 3D scenes from a single image is fundamental to a wide variety of tasks, such as for robotics, motion planning, or augmented reality. Existing works in 3D perception from a single RGB image tend to focus on geometric reconstruction only, or geometric reconstruction with semantic segmentation or instance segmentation. Inspired by 2D panoptic segmentation, we propose to unify the tasks of geometric reconstruction, 3D semantic segmentation, and 3D instance segmentation into the task of panoptic 3D scene reconstruction - from a single RGB image, predicting the complete geometric reconstruction of the scene in the camera frustum of the image, along with semantic and instance segmentations. We thus propose a new approach for holistic 3D scene understanding from a single RGB image which learns to lift and propagate 2D features from an input image to a 3D volumetric scene representation. We demonstrate that this holistic view of joint scene reconstruction, semantic, and instance segmentation is beneficial over treating the tasks independently, thus outperforming alternative approaches.
引用
收藏
页数:12
相关论文
共 50 条
  • [41] Panoptic Lifting for 3D Scene Understanding with Neural Fields
    Siddiqui, Yawar
    Porzi, Lorenzo
    Bulo, Samuel Rota
    Mueller, Norman
    Niessner, Matthias
    Dai, Angela
    Kontschieder, Peter
    [J]. 2023 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2023, : 9043 - 9052
  • [42] EdgeNet: Semantic Scene Completion from a Single RGB-D Image
    Dourado, Aloisio
    De Campos, Teofilo E.
    Kim, Hansung
    Hilton, Adrian
    [J]. 2020 25TH INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION (ICPR), 2021, : 503 - 510
  • [43] PushNet: 3D reconstruction from a single image by pushing
    Ping, Guiju
    Wang, Han
    [J]. NEURAL COMPUTING & APPLICATIONS, 2024, 36 (12): : 6629 - 6641
  • [44] 3D RECONSTRUCTION BASED ON GAT FROM A SINGLE IMAGE
    Yang Dongsheng
    Kuang Ping
    Gu Xiaofeng
    [J]. 2020 17TH INTERNATIONAL COMPUTER CONFERENCE ON WAVELET ACTIVE MEDIA TECHNOLOGY AND INFORMATION PROCESSING (ICCWAMTIP), 2020, : 122 - 125
  • [45] Video supervised for 3D reconstruction from single image
    Zhong, Yijie
    Sun, Zhengxing
    Luo, Shoutong
    Sun, Yunhan
    Wang, Yi
    [J]. MULTIMEDIA TOOLS AND APPLICATIONS, 2022, 81 (11) : 15061 - 15083
  • [46] From Single Image Query to Detailed 3D Reconstruction
    Schonberger, Johannes L.
    Radenovic, Filip
    Chum, Ondrej
    Frahm, Jan-Michael
    [J]. 2015 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2015, : 5126 - 5134
  • [47] 3D corrective nose reconstruction from a single image
    Tang, Yanlong
    Zhang, Yun
    Han, Xiaoguang
    Zhang, Fang-Lue
    Lai, Yu-Kun
    Tong, Ruofeng
    [J]. COMPUTATIONAL VISUAL MEDIA, 2022, 8 (02) : 225 - 237
  • [48] 3D corrective nose reconstruction from a single image
    Yanlong Tang
    Yun Zhang
    Xiaoguang Han
    Fang-Lue Zhang
    Yu-Kun Lai
    Ruofeng Tong
    [J]. Computational Visual Media, 2022, 8 : 225 - 237
  • [49] PushNet: 3D reconstruction from a single image by pushing
    Guiju Ping
    Han Wang
    [J]. Neural Computing and Applications, 2024, 36 : 6629 - 6641
  • [50] A 3D RECONSTRUCTION OF THE HUMAN JAW FROM A SINGLE IMAGE
    Abdelrahim, Aly
    Shalaby, Ahmed
    Elhabian, Shireen
    Graham, James
    Farag, Aly
    [J]. 2013 20TH IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING (ICIP 2013), 2013, : 3622 - 3626