Maximizing Self-supervision from Thermal Image for Effective Self-supervised Learning of Depth and Ego-motion
This work addresses the challenge of generating effective self-supervision from thermal images for depth and ego-motion estimation, which is important for applications in robotics and autonomous systems under challenging conditions, and it is incremental as it builds on existing self-supervised learning methods.
The paper tackled the problem of self-supervised learning of depth and ego-motion from thermal images, which suffer from weak contrast and noise, by proposing a thermal image mapping method that enhances image information while preserving temporal consistency, resulting in outperformed depth and pose results compared to previous state-of-the-art networks without using additional RGB guidance.
Recently, self-supervised learning of depth and ego-motion from thermal images shows strong robustness and reliability under challenging scenarios. However, the inherent thermal image properties such as weak contrast, blurry edges, and noise hinder to generate effective self-supervision from thermal images. Therefore, most research relies on additional self-supervision sources such as well-lit RGB images, generative models, and Lidar information. In this paper, we conduct an in-depth analysis of thermal image characteristics that degenerates self-supervision from thermal images. Based on the analysis, we propose an effective thermal image mapping method that significantly increases image information, such as overall structure, contrast, and details, while preserving temporal consistency. The proposed method shows outperformed depth and pose results than previous state-of-the-art networks without leveraging additional RGB guidance.