Structured Pruning for Efficient Visual Place Recognition

TL;DR

This paper presents a structured pruning method for visual place recognition that reduces memory and latency by over 20% with minimal accuracy loss, optimizing VPR for real-time edge deployment.

Contribution

It introduces a novel structured pruning technique that removes redundancies in VPR architectures and feature embeddings, improving efficiency without significant accuracy compromise.

Findings

Memory usage reduced by 21%

Latency decreased by 16%

Recall@1 accuracy impacted by less than 1%

Abstract

Visual Place Recognition (VPR) is fundamental for the global re-localization of robots and devices, enabling them to recognize previously visited locations based on visual inputs. This capability is crucial for maintaining accurate mapping and localization over large areas. Given that VPR methods need to operate in real-time on embedded systems, it is critical to optimize these systems for minimal resource consumption. While the most efficient VPR approaches employ standard convolutional backbones with fixed descriptor dimensions, these often lead to redundancy in the embedding space as well as in the network architecture. Our work introduces a novel structured pruning method, to not only streamline common VPR architectures but also to strategically remove redundancies within the feature embedding space. This dual focus significantly enhances the efficiency of the system, reducing both map and model memory requirements and decreasing feature extraction and retrieval latencies. Our approach has reduced memory usage and latency by 21% and 16%, respectively, across models, while minimally impacting recall@1 accuracy by less than 1%. This significant improvement enhances real-time applications on edge devices with negligible accuracy loss.