Event-Driven Neuromorphic Vision Enables Energy-Efficient Visual Place Recognition

Researchers developed SpikeVPR, a bio-inspired visual place recognition system using event-based cameras and spiking neural networks that achieves comparable performance to deep networks while using 50x fewer parameters and consuming 30-250x less energy. The neuromorphic approach enables real-time deployment on mobile platforms for autonomous robot navigation.

• →SpikeVPR combines event-based cameras with spiking neural networks for energy-efficient visual place recognition in autonomous robots.
• →The system uses 50 times fewer parameters and consumes 30-250 times less energy than conventional deep networks.
• →Performance matches state-of-the-art deep networks on challenging benchmarks while being robust to illumination and viewpoint changes.
• →EventDilation augmentation strategy enhances robustness to speed and temporal variations in real-world conditions.
• →The approach enables real-time deployment on mobile and neuromorphic platforms for autonomous navigation systems.