DistilVPR: Cross-Modal Knowledge Distillation for Visual Place Recognition
This work addresses the challenge of lightweight deployment for visual place recognition systems, offering a practical solution for resource-constrained applications, though it is incremental as it builds on existing distillation methods.
The paper tackles the problem of high costs and impracticality of multi-modal sensors in visual place recognition by proposing DistilVPR, a knowledge distillation pipeline that enables single-modal systems to learn from cross-modal teachers without extra sensors during inference, achieving state-of-the-art performance compared to other distillation baselines.
The utilization of multi-modal sensor data in visual place recognition (VPR) has demonstrated enhanced performance compared to single-modal counterparts. Nonetheless, integrating additional sensors comes with elevated costs and may not be feasible for systems that demand lightweight operation, thereby impacting the practical deployment of VPR. To address this issue, we resort to knowledge distillation, which empowers single-modal students to learn from cross-modal teachers without introducing additional sensors during inference. Despite the notable advancements achieved by current distillation approaches, the exploration of feature relationships remains an under-explored area. In order to tackle the challenge of cross-modal distillation in VPR, we present DistilVPR, a novel distillation pipeline for VPR. We propose leveraging feature relationships from multiple agents, including self-agents and cross-agents for teacher and student neural networks. Furthermore, we integrate various manifolds, characterized by different space curvatures for exploring feature relationships. This approach enhances the diversity of feature relationships, including Euclidean, spherical, and hyperbolic relationship modules, thereby enhancing the overall representational capacity. The experiments demonstrate that our proposed pipeline achieves state-of-the-art performance compared to other distillation baselines. We also conduct necessary ablation studies to show design effectiveness. The code is released at: https://github.com/sijieaaa/DistilVPR