Self-supervising Fine-grained Region Similarities For Large-scale Image Localization

Abstract

The task of large-scale retrieval-based image localization is to estimate the geographical location of a query image by recognizing its nearest reference images from a city-scale dataset. However, the general public benchmarks only provide noisy GPS labels associated with the training images, which act as weak supervisions for learning image-to-image similarities. Such label noise prevents deep neural networks from learning discriminative features for accurate localization. To tackle this challenge, we propose to self-supervise image-to-region similarities in order to fully explore the potential of difficult positive images alongside their sub-regions. The estimated image-to-region similarities can serve as extra training supervision for improving the network in generations, which could in turn gradually refine the fine-grained similarities to achieve optimal performance. Our proposed self-enhanced image-to-region similarity labels effectively deal with the training bottleneck in the s

Stats

• citations129
• S2 citations—
• github stars0
• HF likes0
• heat score15.85
• arxiv keyge2020self

Related papers

• Data-efficient Large Scale Place Recognition With Graded Similarity Supervision (2023)16.32
• Digging Into Self-supervised Learning Of Feature Descriptors (2021)7.16
• Stochastic Attraction-repulsion Embedding For Large Scale Image Localization (2018)17.57
• Semantic Pose Verification For Outdoor Visual Localization With Self-supervised Contrastive Learning (2022)8.35
• Learning Super-features For Image Retrieval (2022)4.31
• Leveraging Efficientnet And Contrastive Learning For Accurate Global-scale Location Estimation (2021)9.03
• Improving Spatiotemporal Self-supervision By Deep Reinforcement Learning (2018)13.50
• Self-localization From Images With Small Overlap (2016)8.35