Semantic Pose Verification For Outdoor Visual Localization With Self-supervised Contrastive Learning

Abstract

Any city-scale visual localization system has to overcome long-term appearance changes, such as varying illumination conditions or seasonal changes between query and database images. Since semantic content is more robust to such changes, we exploit semantic information to improve visual localization. In our scenario, the database consists of gnomonic views generated from panoramic images (e.g. Google Street View) and query images are collected with a standard field-of-view camera at a different time. To improve localization, we check the semantic similarity between query and database images, which is not trivial since the position and viewpoint of the cameras do not exactly match. To learn similarity, we propose training a CNN in a self-supervised fashion with contrastive learning on a dataset of semantically segmented images. With experiments we showed that this semantic similarity estimation approach works better than measuring the similarity at pixel-level. Finally, we used the sema

Stats

• citations12
• S2 citations—
• github stars0
• HF likes0
• heat score8.35
• arxiv keyorhan2022semantic

Related papers

• Semantic Signatures For Large-scale Visual Localization (2020)5.24
• Domain-invariant Similarity Activation Map Contrastive Learning For Retrieval-based Long-term Visual Localization (2020)13.72
• Self-supervising Fine-grained Region Similarities For Large-scale Image Localization (2020)15.85
• DASGIL: Domain Adaptation For Semantic And Geometric-aware Image-based Localization (2020)13.11
• Training Semantic Descriptors For Image-based Localization (2022)0.00
• Revisiting Contrastive Methods For Unsupervised Learning Of Visual Representations (2021)3.91
• Data-efficient Large Scale Place Recognition With Graded Similarity Supervision (2023)16.32
• Improved Visual Localization Via Graph Smoothing (2019)0.00