Self-Supervised Learning of Camera Ego-Motion From Optical Flow
Chen Xie, Ning Hao, Haodi Yao, Fenghua He · IEEE Transactions on Instrumentation and Measurement · 2025
Estimating the relative pose between two consecutive frames is essential in robotics and computer vision. Recent researches have shown that such ego-motion estimators can be learned from unlabeled monocular videos by joint learning of a pose network and a depth network, which eliminates the need for ground truth data. However, the training speed is slow, training memory is large, and the accuracy of pose prediction is not satisfied. In this paper, we propose a novel and efficient self-supervised learning method to predict the camera ego motion. In light of the uniqueness of flow decomposition we prove, two innovative losses based on optical flow, i.e. epipolar loss and flow decomposition loss, are proposed. The proposed method does not require additional depth network, which leads to faster training process and less training memory. Moreover, to overcome the limitation of performance brought by the presence of moving objects during training, we design a confidence network to automatically localize moving objects and mask the corresponding regions. Experiments across various public datasets and self-collected datasets demonstrate that our ego-motion network achieves state-of-the-art accuracy over long video sequences. Furthermore, our method not only outperforms other self-supervised approaches in terms of accuracy but also surpasses some supervised approaches.