Selectivity Drives Efficiency: Dataset Pruning for Visual Place Recognition
For VPR researchers, this work addresses the inefficiency of large-scale datasets by introducing a pruning method tailored to the pair-based training nature of VPR, outperforming existing DP baselines.
The paper proposes a place-wise dataset pruning framework for visual place recognition (VPR) that treats each place as the pruning unit and uses two metrics (intra-place diversity and inter-place similarity) to select informative data. The method achieves 94.5% R@1 on MSLS-val and 97.0% R@1 on Nordland with NetVLAD while reducing storage and training costs.
Recent visual place recognition (VPR) studies have increasingly relied on large-scale datasets to train more robust and discriminative models. Although this trend significantly improves recognition performance, it also introduces substantial storage and training costs, especially when new architectures or training strategies need to be repeatedly developed and evaluated. Dataset pruning (DP) provides a promising way to improve data efficiency by retaining only informative training data. However, conventional DP methods mainly follow the sample-wise classification paradigm, which overlooks the relation-dependent training nature of VPR, where supervision is typically formed by image pairs rather than independent images. To address this issue, we propose a place-wise dataset pruning framework tailored for VPR. Instead of pruning individual images, our method treats each place as the basic pruning unit and introduces two complementary novel metrics, i.e., intra-place diversity (IPD) and inter-place similarity (IPS), to evaluate the training value of each place. By jointly considering these two metrics, our method ranks all places and constructs a compact yet informative coreset, thereby allowing the pruned dataset to still support the training of robust and discriminative VPR models. Extensive experiments demonstrate that our method consistently outperforms state-of-the-art DP baselines under different pruning ratios while reducing selection and training costs. Moreover, by pruning a merged dataset roughly 3.5$\times$ the size of GSV-Cities to a comparable scale, our coreset maintains highly competitive performance, achieving 94.5\% R@1 on MSLS-val and 97.0\% R@1 on Nordland with only NetVLAD. Codes will be made publicly available.