Simultaneously Learning Corrections and Error Models for Geometry-based Visual Odometry Methods
This addresses accuracy and uncertainty modeling for visual odometry in autonomous driving, representing an incremental improvement through learning-based corrections.
The paper tackles the problem of improving accuracy and uncertainty estimation in classical visual odometry pipelines by using deep learning to learn and compensate for biases, and jointly estimate a full covariance matrix for residual errors. Experiments on autonomous driving image sequences demonstrate concurrent improvements in visual odometry and error estimation.
This paper fosters the idea that deep learning methods can be used to complement classical visual odometry pipelines to improve their accuracy and to associate uncertainty models to their estimations. We show that the biases inherent to the visual odometry process can be faithfully learned and compensated for, and that a learning architecture associated with a probabilistic loss function can jointly estimate a full covariance matrix of the residual errors, defining an error model capturing the heteroscedasticity of the process. Experiments on autonomous driving image sequences assess the possibility to concurrently improve visual odometry and estimate an error associated with its outputs.