Loss it right: Euclidean and Riemannian Metrics in Learning-based Visual Odometry
This work addresses the design of efficient and accurate visual odometry networks for camera motion estimation, but it is incremental as it builds on existing methods like DeepVO.
The paper investigates how different pose representations and loss functions affect the performance of visual odometry networks, finding that using chordal distance as a metric improves generalization and convergence speed.
This paper overviews different pose representations and metric functions in visual odometry (VO) networks. The performance of VO networks heavily relies on how their architecture encodes the information. The choice of pose representation and loss function significantly impacts network convergence and generalization. We investigate these factors in the VO network DeepVO by implementing loss functions based on Euler, quaternion, and chordal distance and analyzing their influence on performance. The results of this study provide insights into how loss functions affect the designing of efficient and accurate VO networks for camera motion estimation. The experiments illustrate that a distance that complies with the mathematical requirements of a metric, such as the chordal distance, provides better generalization and faster convergence. The code for the experiments can be found at https://github.com/remaro-network/Loss_VO_right