Patent
System and Method for Augmenting Channel Characteristics of Audio Recordings
A system to map acoustic characteristics from any low resourced channel audios to larger datasets.
Hi, I'm Shreya.
Welcome to my digital space. I specialize in building complex, scalable multimodal AI solutions across speech, text, audio, and video, with a focus on ethical implementation. Here I share insights and tutorials on the evolving AI landscape, helping others develop powerful tools designed with human values at their core.
Patents and Publications
IEEE SPCOM 2022
Noisy Student Teacher Training with Self Supervised Learning for Children ASR
Proposes a self-supervised learning architecture, achieving 10% relative improvement in Word Error Rate.
IEEE SPCOM 2022
Robustness of DAS Beamformer Over MVDR for Replay Attack Detection On Voice Assistants
Investigates diverse beamforming arithmetic features for replay spoof speech detection in virtual assistants.
