Unsupervised learning for monocular camera motion and 3D scene understanding has gained popularity over traditional methods, which rely on epipolar geometry or non-linear optimization. Notably, deep learning can overcome many issues of monocular vision, such as perceptual aliasing, low-textured areas, scale drift, and degenerate motions. In addition, concerning supervised learning, we can fully leverage video stream data without the need for depth or motion labels. However, in this work, we note that rotational motion can limit the accuracy of the unsupervised pose networks more than the translational component. Therefore, we present RAUM-VO, an approach based on a model-free epipolar constraint for frame-to-frame motion estimation (F2F) to adjust the rotation during training and online inference. To this end, we match 2D keypoints between consecutive frames using pre-trained deep networks, Superpoint and Superglue, while training a network for depth and pose estimation using an unsupervised training protocol. Then, we adjust the predicted rotation with the motion estimated by F2F using the 2D matches and initializing the solver with the pose network prediction. Ultimately, RAUM-VO shows a considerable accuracy improvement compared to other unsupervised pose networks on the KITTI dataset, while reducing the complexity of other hybrid or traditional approaches and achieving comparable state-of-the-art results.
Keywords: deep learning; depth estimation; unsupervised learning; visual odometry.