arrow
返回

Structured Pruning for Efficient Visual Place Recognition

delete2024-01-01
delete0
delete
OA
AI
O
Oliver Grainge
M
Michael Milford
I
Indu Bodala
S
Sarvapali D. Ramchurn
S
Shoaib Ehsan
DOI:10.1109/lra.2024.3523222delete
delete原文链接
delete分享
delete收藏
查看原文
摘要

摘要

En 中文
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.
AI总结

AI总结

对已上传原文的论文进行重点信息的提取,主要内容包括:简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。

期刊

I
IEEE Robotics and Automation Letters
IF:
5.3
论文数:
3.4K
被引数:
4.0W

机构

U
University of Southampton
学者数:
1.5K
论文数: 779
被引数: 2.2W
Q
Queensland University of Technology
学者数:
969
论文数: 434
被引数: 560
引用论文

引用论文

Design Space Exploration of Low-Bit Quantized Neural Networks for Visual Place Recognition
err
IF0
err2024-06-01
err0
PREAI
errOliver Grainge; Michael Milford; Indu Bodala; Sarvapali D. Ramchurn; Shoaib Ehsan
err分享
err收藏
err分享
err收藏
Deep learning features at scale for visual place recognition
err2017-05-01
err0
errOAAI
errZetao Chen; Adam Jacobson; Niko Sunderhauf; Ben Upcroft; Lingqiao Liu; Chunhua Shen; Ian Reid; Michael Milford
err分享
err收藏
err分享
err收藏
err分享
err收藏
Mapillary Street-Level Sequences: A Dataset for Lifelong Place Recognition
err2020-06-01
err0
errOAAI
errFrederik Warburg; Soren Hauberg; Manuel Lopez-Antequera; Pau Gargallo; Yubin Kuang; Javier Civera
err分享
err收藏
Binary Neural Networks for Memory-Efficient and Effective Visual Place Recognition in Changing Environments
err2022-08-01
err11
errOAAI
errFerrarini, Bruno; Milford, Michael J.; McDonald-Maier, Klaus D.; Ehsan, Shoaib
err分享
err收藏
学者 查看更多内容