A10.13uJ/classification 2-channel Deep Neural Network-based SoC for Emotion Detection of Autistic Children
Abdul Rehman Aslam, Talha Iqbal, Mahnoor Aftab, Wala Saadeh, Muhammad Awais Bin Altaf
Abstract
An EEG-based noninvasive neuro-feedback SoC for emotion classification of Autistic children is presented. The AFE comprises two entirely shared EEG-channels using sampling capacitors to reduce the area by 30% and achieve an overall integrated input-referred noise of 0.55μ VRMS with cross-talk of - 79dB. The 4-layers Deep Neural Network (DNN) classifier is integrated on-sensor to classify (4 emotions) with >85% accuracy. The 16mm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> SoC in 0.18um CMOS consumes 10.13μJ/classification for 2 channels.
Topics & Concepts
Computer scienceElectroencephalographyClassifier (UML)Artificial neural networkChannel (broadcasting)Artificial intelligenceAutismSpeech recognitionPattern recognition (psychology)PsychologyTelecommunicationsNeuroscienceDevelopmental psychologyAnalog and Mixed-Signal Circuit DesignEEG and Brain-Computer InterfacesNeural dynamics and brain function