A neurally inspired pattern recognition approach with latency-phase encoding and precise-spike-driven rule in spiking neural network

Recent years, several advances have been made in spiking neural networks (SNNs). In this paper, we introduce a novel approach that uses both the latency-phase encoding and the precise-spike-driven (PSD) learning rule for pattern recognition. The empirical results show that our proposed approach can perform quite competitive in comparison to other commonly used encoding and learning strategies in SNNs, when evaluating on the gray OCR dataset and the Address-Event-Representation (AER) motion dataset for image classification.