Differentiable Mapping Networks: Learning Structured Map Representations for Sparse Visual Localization

TL;DR

This paper introduces the Differentiable Mapping Network (DMN), a neural network architecture that jointly learns spatially structured map representations and visual localization, improving performance especially in large, data-scarce environments.

Contribution

The paper presents a novel end-to-end differentiable neural network architecture for structured mapping and localization, combining spatial structure with gradient-based learning.

Findings

DMN effectively learns map representations for visual localization.

Spatial structure benefits increase with environment size and fewer training data.

DMN outperforms baseline methods in simulated and real-world datasets.

Abstract

Mapping and localization, preferably from a small number of observations, are fundamental tasks in robotics. We address these tasks by combining spatial structure (differentiable mapping) and end-to-end learning in a novel neural network architecture: the Differentiable Mapping Network (DMN). The DMN constructs a spatially structured view-embedding map and uses it for subsequent visual localization with a particle filter. Since the DMN architecture is end-to-end differentiable, we can jointly learn the map representation and localization using gradient descent. We apply the DMN to sparse visual localization, where a robot needs to localize in a new environment with respect to a small number of images from known viewpoints. We evaluate the DMN using simulated environments and a challenging real-world Street View dataset. We find that the DMN learns effective map representations for visual localization. The benefit of spatial structure increases with larger environments, more viewpoints for mapping, and when training data is scarce. Project website: http://sites.google.com/view/differentiable-mapping