Domain-invariant Similarity Activation Map Contrastive Learning For Retrieval-based Long-term Visual Localization

Abstract

Visual localization is a crucial component in the application of mobile robot and autonomous driving. Image retrieval is an efficient and effective technique in image-based localization methods. Due to the drastic variability of environmental conditions, e.g. illumination, seasonal and weather changes, retrieval-based visual localization is severely affected and becomes a challenging problem. In this work, a general architecture is first formulated probabilistically to extract domain invariant feature through multi-domain image translation. And then a novel gradient-weighted similarity activation mapping loss (Grad-SAM) is incorporated for finer localization with high accuracy. We also propose a new adaptive triplet loss to boost the contrastive learning of the embedding in a self-supervised manner. The final coarse-to-fine image retrieval pipeline is implemented as the sequential combination of models without and with Grad-SAM loss. Extensive experiments have been conducted to validat

Stats

• citations27
• S2 citations—
• github stars26
• HF likes0
• heat score13.72
• arxiv keyhu2020domain

Related papers

• Retrieval-based Localization Based On Domain-invariant Feature Learning Under Changing Environments (2019)9.92
• DASGIL: Domain Adaptation For Semantic And Geometric-aware Image-based Localization (2020)13.11
• Learning Condition Invariant Features For Retrieval-based Localization From 1M Images (2020)0.00
• Semantic Pose Verification For Outdoor Visual Localization With Self-supervised Contrastive Learning (2022)8.35
• Night-to-day Image Translation For Retrieval-based Localization (2018)16.88
• Investigating The Role Of Image Retrieval For Visual Localization -- An Exhaustive Benchmark (2022)16.58
• Benchmarking Image Retrieval For Visual Localization (2020)17.78
• Leveraging Efficientnet And Contrastive Learning For Accurate Global-scale Location Estimation (2021)9.03