StructVPR++: Distill Structural and Semantic Knowledge with Weighting Samples for Visual Place Recognition

2025-03-09

AI Summary

Visual Place Recognition (VPR) faces dual challenges in autonomous driving and robotics: insufficient semantic discriminability of global features and high computational overhead in re-ranking. This paper proposes an end-to-end, RGB-only global feature learning framework to bridge the accuracy–efficiency gap. First, we introduce a novel label-aware feature disentanglement mechanism that enables explicit semantic alignment at inference time—without requiring segmentation masks. Second, we design segmentation-guided knowledge distillation and sample-weighted loss to dynamically suppress noisy image pairs and strengthen reliable supervision signals. Evaluated on four standard benchmarks, our method achieves 5–23% improvements in Recall@1 over state-of-the-art global-feature-based approaches, matching the performance of two-stage methods while enabling real-time, single-frame inference.

Technology Category

• Computer Vision: Vision for Robotics & Autonomous Driving
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Machine Learning: Adversarial Learning & Robustness

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Visual place recognition is a challenging task for autonomous driving and robotics, which is usually considered as an image retrieval problem. A commonly used two-stage strategy involves global retrieval followed by re-ranking using patch-level descriptors. Most deep learning-based methods in an end-to-end manner cannot extract global features with sufficient semantic information from RGB images. In contrast, re-ranking can utilize more explicit structural and semantic information in one-to-one matching process, but it is time-consuming. To bridge the gap between global retrieval and re-ranking and achieve a good trade-off between accuracy and efficiency, we propose StructVPR++, a framework that embeds structural and semantic knowledge into RGB global representations via segmentation-guided distillation. Our key innovation lies in decoupling label-specific features from global descriptors, enabling explicit semantic alignment between image pairs without requiring segmentation during deployment. Furthermore, we introduce a sample-wise weighted distillation strategy that prioritizes reliable training pairs while suppressing noisy ones. Experiments on four benchmarks demonstrate that StructVPR++ surpasses state-of-the-art global methods by 5-23% in Recall@1 and even outperforms many two-stage approaches, achieving real-time efficiency with a single RGB input.

Problem

• Bridges gap between global retrieval and re-ranking for visual place recognition
• Embeds structural and semantic knowledge into RGB global representations
• Improves accuracy and efficiency in autonomous driving and robotics

Innovation

• Segmentation-guided distillation for knowledge embedding
• Decoupling label-specific features for semantic alignment
• Sample-wise weighted distillation to prioritize reliable pairs