Harshavardhana Gowda

My work focuses on developing neural interfaces using surface electromyogram signals. One branch of my work attempts to develop hand-gesture based interfaces for establishing seamless human-computer interaction such as hand gesture and hand writing decoding using sEMG signals. Another arm of my work attends to synthesis of speech using sEMG signals from orofacial muscles.
Given the necessity for personalization and limited availability of data, my work focuses on developing efficient geometry-inspired neural architectures for processing multivariate time series and audio signals.