Event-Based localization with Spiking Neural Networks: How Much is needed ?
Abstract
The aim of this project is to develop bio-inspired and energy efficient computational models for navigation in space. These models should be able to achieve Visual Place Recognition (VPR), Odometry and finally Visual Simultaneous Localization and Mapping (V-SLAM) from spiking neural networks (SNNs) processing data collected by an event-based camera.
Origin | Files produced by the author(s) |
---|---|
Licence |