Global Proxy-based Hard Mining for Visual Place Recognition

TL;DR

This paper introduces a global proxy-based hard mining technique for visual place recognition that improves training efficiency and accuracy by using proxy representations to sample hard examples, achieving state-of-the-art results.

Contribution

The authors propose an end-to-end trainable proxy-based sampling method that enhances deep visual place recognition training without significant computational overhead.

Findings

Achieves state-of-the-art performance on multiple benchmarks.

Provides over 100% relative improvement on the Nordland dataset.

Enables effective hard example mining with minimal additional cost.

Abstract

Learning deep representations for visual place recognition is commonly performed using pairwise or triple loss functions that highly depend on the hardness of the examples sampled at each training iteration. Existing techniques address this by using computationally and memory expensive offline hard mining, which consists of identifying, at each iteration, the hardest samples from the training set. In this paper we introduce a new technique that performs global hard mini-batch sampling based on proxies. To do so, we add a new end-to-end trainable branch to the network, which generates efficient place descriptors (one proxy for each place). These proxy representations are thus used to construct a global index that encompasses the similarities between all places in the dataset, allowing for highly informative mini-batch sampling at each training iteration. Our method can be used in combination with all existing pairwise and triplet loss functions with negligible additional memory and computation cost. We run extensive ablation studies and show that our technique brings new state-of-the-art performance on multiple large-scale benchmarks such as Pittsburgh, Mapillary-SLS and SPED. In particular, our method provides more than 100% relative improvement on the challenging Nordland dataset. Our code is available at https://github.com/amaralibey/GPM