One Channel to Rule Them All: Rethinking Input Representation for Visual Place Recognition

2026-05-30

AI Summary

This study challenges the prevailing assumption in visual place recognition that RGB inputs are essential, presenting a systematic evaluation of grayscale versus RGB across diverse model architectures, training strategies, and benchmark datasets. The experiments demonstrate that, in most scenarios, single-channel grayscale images not only serve as an effective substitute for RGB but often yield superior performance—particularly under significant appearance variations such as changes in illumination or season. Using mainstream models like MixVPR, grayscale-trained variants achieve an average Recall@1 of 82.4%, outperforming their RGB counterparts at 81.2%. Moreover, lightweight grayscale models reduce parameters by 60% while still surpassing heavier RGB-based models, offering compelling advantages in storage efficiency and bandwidth requirements, thereby opening new avenues for efficient deployment.

Abstract

Visual Place Recognition (VPR) is fundamental to long-term robot localization and SLAM, yet current systems overwhelmingly rely on RGB input, implicitly assuming color is necessary for global place recognition. We challenge this assumption, investigating the role of chromatic information across training regimes, model architectures and standard benchmarks under real-world appearance variation. We find that grayscale matches RGB performance generally and outperforms it under severe appearance shifts where color invariance is insufficiently learned, while color provides meaningful gains only where persistent and discriminative chromatic cues are present. Across selected benchmarks, a fully gray-trained MixVPR model achieves an average 82.4% Recall@1 compared to 81.2% for its RGB counterpart. In some cases, lightweight grayscale variants with 60% fewer parameters can outperform heavier RGB models. Grayscale further offers practical advantages in storage, bandwidth and alignment with resource-constrained systems. We conclude that for global VPR where scenes vary across illumination, weather, season and setting, color contributes minimally, and grayscale alone is sufficient for reliable place recognition.

Problem

• Visual Place Recognition
• RGB input
• grayscale representation
• color invariance
• appearance variation

Innovation

• grayscale representation
• visual place recognition
• color invariance
• resource-efficient VPR
• appearance variation