Accession Number : AD1003123


Title :   Memristive Computational Architecture of an Echo State Network for Real-Time Speech Emotion Recognition


Descriptive Note : Conference Paper


Corporate Author : Air Force Research Laboratory/RITB Rome United States


Personal Author(s) : Saleh,Qutaiba ; Merkel,Cory ; Kudithipudi,Dhireesha ; Wysocki,Bryant


Full Text : https://apps.dtic.mil/dtic/tr/fulltext/u2/1003123.pdf


Report Date : 28 May 2015


Pagination or Media Count : 6


Abstract : Echo state networks (ESNs) provide an efficient classification technique for spatiotemporal signals. The feedback connections in the ESN topology enable feature extraction of both spatial and temporal components in time series data. This property has been used in several application domains such as image and video analysis, anomaly detection, andspeech recognition. In this research, a hardware architecture was explored for realizing ESN efficiently in power constrained devices. Specifically, a scalable computational architecture applied to speech-emotion recognition was proposed. Two different topologies were explored, with memristive synapses. The simulation results are promising with a classification accuracy of approximately equals 96% for two distinct emotion statuses.


Descriptors :   artificial neural networks , signal processing , simulations , circuits , speech


Distribution Statement : APPROVED FOR PUBLIC RELEASE