To Match or Not to Match: Revisiting Image Matching for Reliable Visual Place Recognition
This work addresses the reliability of VPR systems for robotics and autonomous navigation by shifting the paradigm from mandatory re-ranking to adaptive verification, though it is incremental as it builds on existing retrieval and matching methods.
The paper tackles the problem of whether to use image matching for re-ranking in Visual Place Recognition (VPR), showing that modern retrieval systems often degrade with re-ranking due to dataset saturation, and proposes using inlier counts from image matching as a verification step to predict when re-ranking is beneficial.
Visual Place Recognition (VPR) is a critical task in computer vision, traditionally enhanced by re-ranking retrieval results with image matching. However, recent advancements in VPR methods have significantly improved performance, challenging the necessity of re-ranking. In this work, we show that modern retrieval systems often reach a point where re-ranking can degrade results, as current VPR datasets are largely saturated. We propose using image matching as a verification step to assess retrieval confidence, demonstrating that inlier counts can reliably predict when re-ranking is beneficial. Our findings shift the paradigm of retrieval pipelines, offering insights for more robust and adaptive VPR systems. The code is available at https://github.com/FarInHeight/To-Match-or-Not-to-Match.