Prior Guided Dropout for Robust Visual Localization in Dynamic Environments

Zhaoyang Huang, Yan Xu, Jianping Shi, Xiaowei Zhou, Hujun Bao, Guofeng Zhang; The IEEE International Conference on Computer Vision (ICCV), 2019, pp. 2791-2800

Abstract


Camera localization from monocular images has been a long-standing problem, but its robustness in dynamic environments is still not adequately addressed. Compared with classic geometric approaches, modern CNN-based methods (e.g. PoseNet) have manifested the reliability against illumination or viewpoint variations, but they still have the following limitations. First, foreground moving objects are not explicitly handled, which results in poor performance and instability in dynamic environments. Second, the output for each image is a point estimate without uncertainty quantification. In this paper, we propose a framework which can be generally applied to existing CNN-based pose regressors to improve their robustness in dynamic environments. The key idea is a prior guided dropout module coupled with a self-attention module which can guide CNNs to ignore foreground objects during both training and inference. Additionally, the dropout module enables the pose regressor to output multiple hypotheses from which the uncertainty of pose estimates can be quantified and leveraged in the following uncertainty-aware pose-graph optimization to improve the robustness further. We achieve an average accuracy of 9.98m/3.63deg on RobotCar dataset, which outperforms the state-of-the-art method by 62.97%/47.08%. The source code of our implementation is available at https://github.com/zju3dv/RVL-dynamic.

Related Material


[pdf] [supp]
[bibtex]
@InProceedings{Huang_2019_ICCV,
author = {Huang, Zhaoyang and Xu, Yan and Shi, Jianping and Zhou, Xiaowei and Bao, Hujun and Zhang, Guofeng},
title = {Prior Guided Dropout for Robust Visual Localization in Dynamic Environments},
booktitle = {The IEEE International Conference on Computer Vision (ICCV)},
month = {October},
year = {2019}
}