Adversarial Learning Of Hard Positives For Place Recognition

Abstract

Image retrieval methods for place recognition learn global image descriptors that are used for fetching geo-tagged images at inference time. Recent works have suggested employing weak and self-supervision for mining hard positives and hard negatives in order to improve localization accuracy and robustness to visibility changes (e.g. in illumination or view point). However, generating hard positives, which is essential for obtaining robustness, is still limited to hard-coded or global augmentations. In this work we propose an adversarial method to guide the creation of hard positives for training image retrieval networks. Our method learns local and global augmentation policies which will increase the training loss, while the image retrieval network is forced to learn more powerful features for discriminating increasingly difficult examples. This approach allows the image retrieval network to generalize beyond the hard examples presented in the data and learn features that are robust to

Stats

• citations1
• S2 citations—
• github stars0
• HF likes0
• heat score2.26
• arxiv keyfang2022adversarial

Related papers

• Adversarial Training For Adverse Conditions: Robust Metric Localisation Using Appearance Transfer (2018)14.43
• Digging Into Self-supervised Learning Of Feature Descriptors (2021)7.16
• Are Local Features All You Need For Cross-domain Visual Place Recognition? (2023)13.80
• Learning Condition Invariant Features For Retrieval-based Localization From 1M Images (2020)0.00
• Logg3d-net: Locally Guided Global Descriptor Learning For 3D Place Recognition (2021)19.02
• VSE++: Improving Visual-semantic Embeddings With Hard Negatives (2017)0.00
• Focus On Local: Finding Reliable Discriminative Regions For Visual Place Recognition (2025)10.70
• Eigenplaces: Training Viewpoint Robust Models For Visual Place Recognition (2023)15.46