Perceptual Loss for Robust Unsupervised Homography Estimation
This addresses robustness issues in homography estimation for computer vision tasks, but it is incremental as it builds on existing unsupervised methods with a novel loss.
The paper tackles the problem of robust homography estimation under illumination and viewpoint changes by proposing a bidirectional implicit Homography Estimation (biHomE) loss for unsupervised learning, achieving state-of-the-art performance on synthetic COCO dataset that is comparable to supervised methods.
Homography estimation is often an indispensable step in many computer vision tasks. The existing approaches, however, are not robust to illumination and/or larger viewpoint changes. In this paper, we propose bidirectional implicit Homography Estimation (biHomE) loss for unsupervised homography estimation. biHomE minimizes the distance in the feature space between the warped image from the source viewpoint and the corresponding image from the target viewpoint. Since we use a fixed pre-trained feature extractor and the only learnable component of our framework is the homography network, we effectively decouple the homography estimation from representation learning. We use an additional photometric distortion step in the synthetic COCO dataset generation to better represent the illumination variation of the real-world scenarios. We show that biHomE achieves state-of-the-art performance on synthetic COCO dataset, which is also comparable or better compared to supervised approaches. Furthermore, the empirical results demonstrate the robustness of our approach to illumination variation compared to existing methods.