Spike‐Enabled Audio Learning in Multilevel Synaptic Memristor Array‐Based Spiking Neural Network. (18th October 2021)

• Record Type: Journal Article
• Authors: Wu, Xulei; Dang, Bingjie; Wang, Hong; Wu, Xiulong; Yang, Yuchao
• Is Part Of: Advanced intelligent systems. Volume 4:Number 3(2022)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2021-10-18
• Subjects: long-term plasticity -- memristors -- speech recognition -- spiking neural networks; Artificial intelligence -- Periodicals; Robotics -- Periodicals; Control theory -- Periodicals; 006.3
• DOI: 10.1002/aisy.202100151
• Languages: English
• ISSNs: 2640-4567

Abstract

Abstract : Speech recognition involves the ability to learn the audios which are closely related to event sequence. Although speech recognition has been widely implemented in software neural networks, a hardware implementation based on energy efficient computing architecture is still missing. Herein, W/MgO/SiO2 /Mo memristor arrays with multilevel resistance states are fabricated, where the weights of the artificial synapses in the memristor array can be tuned precisely by voltage pulses. Based on the array, speech recognition in memristive spiking neural networks (SNNs) with improved supervised tempotron algorithm on Texas Instruments digit sequences (TIDIGITS) dataset is conducted, demonstrating software‐comparable accuracy for speech recognition in the memristive SNN. It is envisioned that such memristive SNNs can pave the way to building a bioinspired spike‐based neuromorphic system for audio learning. Abstract : Herein, a W/MgO/SiO2 /Mo memristor array with multilevel resistance states, where the weights of the memristor array can be tuned precisely by voltage pulses, is reported. Based on the array, speech recognition in memristive spiking neural network with improved supervised tempotron algorithm on TIDIGITS dataset is conducted, demonstrating software‐comparable accuracy for speech recognition in the memristive spiking neural network.