Bootstrapped Self-Supervised Training with Monocular Video for Semantic Segmentation and Depth Estimation
This addresses the need for autonomous learning in robotics to refine pre-set knowledge, though it is incremental as it builds on existing self-supervised techniques.
The paper tackles the problem of improving semantic segmentation and depth estimation for robots by proposing a bootstrapped self-supervised training method that uses temporal consistency in monocular video, showing it enhances a state-of-the-art semantic segmentation network and outperforms pure supervised or self-supervised training for depth estimation.
For a robot deployed in the world, it is desirable to have the ability of autonomous learning to improve its initial pre-set knowledge. We formalize this as a bootstrapped self-supervised learning problem where a system is initially bootstrapped with supervised training on a labeled dataset and we look for a self-supervised training method that can subsequently improve the system over the supervised training baseline using only unlabeled data. In this work, we leverage temporal consistency between frames in monocular video to perform this bootstrapped self-supervised training. We show that a well-trained state-of-the-art semantic segmentation network can be further improved through our method. In addition, we show that the bootstrapped self-supervised training framework can help a network learn depth estimation better than pure supervised training or self-supervised training.