Information-Theoretic Odometry Learning
This work addresses odometry estimation for robotics and virtual reality applications, offering a novel theoretical framework that is incremental in applying information theory to an existing problem.
The paper tackles the problem of odometry estimation for robotics and vision tasks by proposing an information-theoretic framework that optimizes a variational information bottleneck to remove pose-irrelevant information, resulting in a method that provides performance guarantees, practical design guidance, and natural uncertainty measures, with effectiveness demonstrated on two datasets.
In this paper, we propose a unified information theoretic framework for learning-motivated methods aimed at odometry estimation, a crucial component of many robotics and vision tasks such as navigation and virtual reality where relative camera poses are required in real time. We formulate this problem as optimizing a variational information bottleneck objective function, which eliminates pose-irrelevant information from the latent representation. The proposed framework provides an elegant tool for performance evaluation and understanding in information-theoretic language. Specifically, we bound the generalization errors of the deep information bottleneck framework and the predictability of the latent representation. These provide not only a performance guarantee but also practical guidance for model design, sample collection, and sensor selection. Furthermore, the stochastic latent representation provides a natural uncertainty measure without the needs for extra structures or computations. Experiments on two well-known odometry datasets demonstrate the effectiveness of our method.