معرفی
Sarthak Yadav is a PhD Fellow at the Department of Electronic Systems, Technical Faculty of IT and Design, Aalborg University. His research focuses on self-supervised learning for general-purpose audio representations, with affiliations to the Signals and Decoding collaboratory at the Pioneer Center for Artificial Intelligence in Copenhagen.
- Education: MSc(R) in Computer Science (University of Glasgow, 2022), Bachelor of Technology in Computer Science and Engineering (APJ Abdul Kalam Technological University, 2017)
Yadav specializes in training large-scale deep neural networks on unlabelled audio data. His work involves developing advanced architectures like state spaces (Mamba), xLSTMs, and masked autoencoders with multi-window attention mechanisms to improve audio representation learning. These methods aim to enhance sequence modeling, emotion recognition, and time-frequency analysis.
Recent publications (2024) highlight applications of selective state spaces, xLSTMs, and multi-window attention frameworks in audio processing. Earlier works focus on masked autoencoders and speech emotion recognition techniques.
Yadav presented his research at major conferences including Interspeech 2024 and ICASSP 2023. He is actively engaged in collaborative projects under supervisors from Aalborg University and the Pioneer Center for Artificial Intelligence.
