EGA-Depth: Efficient Guided Attention for Self-Supervised Multi-Camera Depth Estimation

被引:5
|
作者
Shi, Yunxiao [1 ]
Cai, Hong [1 ]
Ansari, Amin [2 ]
Porikli, Fatih [1 ]
机构
[1] Qualcomm AI Res, San Diego, CA 92121 USA
[2] Qualcomm Technol Inc, San Diego, CA USA
关键词
D O I
10.1109/CVPRW59228.2023.00017
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
The ubiquitous multi-camera setup on modern autonomous vehicles provides an opportunity to construct surround-view depth. Existing methods, however, either perform independent monocular depth estimations on each camera or rely on computationally heavy self attention mechanisms. In this paper, we propose a novel guided attention architecture, EGA-Depth, which can improve both the efficiency and accuracy of self-supervised multi-camera depth estimation. More specifically, for each camera, we use its perspective view as the query to cross-reference its neighboring views to derive informative features for this camera view. This allows the model to perform attention only across views with considerable overlaps and avoid the costly computations of standard self-attention. Given its efficiency, EGA-Depth enables us to exploit higher-resolution visual features, leading to improved accuracy. Furthermore, EGA-Depth can incorporate more frames from previous time steps as it scales linearly w.r.t. the number of views and frames. Extensive experiments on two challenging autonomous driving benchmarks nuScenes and DDAD demonstrate the efficacy of our proposed EGA-Depth and show that it achieves the new state-of-the-art in self-supervised multi-camera depth estimation.
引用
收藏
页码:119 / 129
页数:11
相关论文
共 50 条
  • [21] Self-supervised monocular depth estimation in fog
    Tao, Bo
    Hu, Jiaxin
    Jiang, Du
    Li, Gongfa
    Chen, Baojia
    Qian, Xinbo
    OPTICAL ENGINEERING, 2023, 62 (03)
  • [22] On the uncertainty of self-supervised monocular depth estimation
    Poggi, Matteo
    Aleotti, Filippo
    Tosi, Fabio
    Mattoccia, Stefano
    2020 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2020, : 3224 - 3234
  • [23] Revisiting Self-supervised Monocular Depth Estimation
    Kim, Ue-Hwan
    Lee, Gyeong-Min
    Kim, Jong-Hwan
    ROBOT INTELLIGENCE TECHNOLOGY AND APPLICATIONS 6, 2022, 429 : 336 - 350
  • [24] Attention meets Geometry: Geometry Guided Spatial-Temporal Attention for Consistent Self-Supervised Monocular Depth Estimation
    Ruhkamp, Patrick
    Gao, Daoyi
    Chen, Hanzhi
    Navab, Nassir
    Busam, Beniamin
    2021 INTERNATIONAL CONFERENCE ON 3D VISION (3DV 2021), 2021, : 837 - 847
  • [25] Multi-resolution distillation for self-supervised monocular depth estimation
    Lee, Sebin
    Im, Woobin
    Yoon, Sung-Eui
    PATTERN RECOGNITION LETTERS, 2023, 176 : 215 - 222
  • [26] Multi-resolution distillation for self-supervised monocular depth estimation
    Lee, Sebin
    Im, Woobin
    Yoon, Sung-Eui
    PATTERN RECOGNITION LETTERS, 2023, 176
  • [27] ADAADepth: Adapting Data Augmentation and Attention for Self-Supervised Monocular Depth Estimation
    Kaushik, Vinay
    Jindgar, Kartik
    Lall, Brejesh
    IEEE ROBOTICS AND AUTOMATION LETTERS, 2021, 6 (04) : 7791 - 7798
  • [28] Joint Soft-Hard Attention for Self-Supervised Monocular Depth Estimation
    Fan, Chao
    Yin, Zhenyu
    Xu, Fulong
    Chai, Anying
    Zhang, Feiqing
    SENSORS, 2021, 21 (21)
  • [29] A multi-camera dataset for depth estimation in an indoor scenario
    Marin, Giulio
    Agresti, Gianluca
    Minto, Ludovico
    Zanuttigh, Pietro
    DATA IN BRIEF, 2019, 27
  • [30] Depth360: Self-supervised Learning for Monocular Depth Estimation using Learnable Camera Distortion Model
    Hirose, Noriaki
    Tahara, Kosuke
    2022 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS), 2022, : 317 - 324