Adaptive Continuous Visual Odometry from RGB-D Images

TL;DR

This paper introduces an adaptive continuous visual odometry method for RGB-D cameras that automatically learns optimal hyperparameters online, improving performance over fixed-parameter approaches.

Contribution

It proposes an online hyperparameter learning approach using gradient descent for adaptive continuous visual odometry, enhancing robustness and accuracy.

Findings

Outperforms original framework and state-of-the-art methods on RGB-D benchmarks.

Automatically adapts length-scale hyperparameters during operation.

Provides publicly available software implementation.

Abstract

In this paper, we extend the recently developed continuous visual odometry framework for RGB-D cameras to an adaptive framework via online hyperparameter learning. We focus on the case of isotropic kernels with a scalar as the length-scale. In practice and as expected, the length-scale has remarkable impacts on the performance of the original framework. Previously it was handled using a fixed set of conditions within the solver to reduce the length-scale as the algorithm reaches a local minimum. We automate this process by a greedy gradient descent step at each iteration to find the next-best length-scale. Furthermore, to handle failure cases in the gradient descent step where the gradient is not well-behaved, such as the absence of structure or texture in the scene, we use a search interval for the length-scale and guide it gradually toward the smaller values. This latter strategy reverts the adaptive framework to the original setup. The experimental evaluations using publicly available RGB-D benchmarks show the proposed adaptive continuous visual odometry outperforms the original framework and the current state-of-the-art. We also make the software for the developed algorithm publicly available.