Talk details
Title: DeMoAiR: Deep Learning based Multi-modal Airwriting Recognition for Smart Wearables
Speaker: Ayush Tripathi (PhD, IIT Delhi & Postdoc, Harvard Medical School), Assistant Professor, IIT Indore
Date: 11th October 2025
Time: 11:00 AM (IST)
Abstract:
Airwriting is the process of writing letters in free space with natural finger movements. It offers an intuitive alternative for human computer interaction applications, requiring no new gestures to be learnt for being able to use such a system. In this talk, we explore a multi-modal approach to airwriting recognition using brain, muscle, and motion signals, captured via Electroencephalography (EEG), Electromyography (EMG), and wrist-worn Inertial Measurement Unit (IMU). We present insights into the feasibility of each modality and demonstrate a real-time recognition system. Airwriting holds strong potential as a user-friendly, hands-free input method for next-generation smart wearables and human-computer interaction.
Speaker’s bio
Ayush Tripathi is an Assistant Professor in the Department of Electrical Engineering at IIT Indore, where he is establishing the COmputational Signal INformatics and Exploration (COSINE) Lab, focusing on biomedical signal processing and AI applications in healthcare. His research spans brain-computer and human-machine interfaces, EEG analysis, automated sleep analysis, and pathological speech processing. He earned his Ph.D. from IIT Delhi, was a Postdoctoral Fellow at Harvard Medical School, and previously worked at TCS Research & Innovation after completing his B.Tech. from VNIT Nagpur.