Balancing Performance and Efficiency in Zero-shot Robotic Navigation

TL;DR

This paper evaluates various vision-language models and modules for zero-shot robotic navigation, proposing an optimized solution that improves success rate and reduces memory usage in resource-limited settings.

Contribution

It introduces an optimized approach that balances navigation success and computational efficiency, outperforming baseline models with less memory.

Findings

Success rate increased by 1.55% over baseline

Achieved 2.3 times less video memory usage

Provided insights into resource-efficient deployment strategies

Abstract

We present an optimization study of the Vision-Language Frontier Maps (VLFM) applied to the Object Goal Navigation task in robotics. Our work evaluates the efficiency and performance of various vision-language models, object detectors, segmentation models, and multi-modal comprehension and Visual Question Answering modules. Using the $\textit{val-mini}$ and $\textit{val}$ splits of Habitat-Matterport 3D dataset, we conduct experiments on a desktop with limited VRAM. We propose a solution that achieves a higher success rate (+1.55%) improving over the VLFM BLIP-2 baseline without substantial success-weighted path length loss while requiring $\textbf{2.3 times}$ less video memory. Our findings provide insights into balancing model performance and computational efficiency, suggesting effective deployment strategies for resource-limited environments.