I’m Shreekantha (Shree). I am pursuing Master of Science by Research at the International Institute of Information Technology, Bangalore (IIIT-B), India.
My thesis is on “Attention-based end-to-end frameworks for Automatic Speech Recognition”.
I am trying to bring in some classical speech knowledge to the purely data-driven models of recent years. Through this exercise, we are hoping to blend the pure data-driven architectures with speech knowledge, leading to a reduction in model complexity, faster training/inference and deeper insights into speech recognition. I am in the final stages of my thesis and planning to submit in a couple of months. Other problems I am working on for my thesis include KWS using attention, a pre-training method for end-to-end ASR, better training strategies for encoder-attention-decoder models, interpretability, explainability of end-to-end ASR models.
Most of my thesis work is done with kaldi, ESPnet and some custom code I have written in PyTorch and TensorFlow.
At IIIT-Bangalore, I am also involved in different labs like the E-health research centre (EHRC) where we develop rehabilitation robotics applications, the centre for Machine Intelligence and Robotics (MINRO), where we study multi-lingual applications of Speech and Language technologies in the domain of e-governance. I have been a Graduate Teaching Assistant for courses including
- Deep Learning for Automatic Speech Recognition
- Automatic Speech Recognition
- Introduction to Robotics
which involves delivering talks, tutorials, and interaction with course students.
In my time at IIIT-B, I have also mentored many Master’s students in projects and thesis in Speech Recognition, Machine Learning and Signal Processing. I am also part of the AI reading club at IIIT-B and have delivered talks on many occasions.
Recently, I have also been selected as an Intel AI Academy Student Ambassador. We are trying to build a small-footprint ASR applications (keyword spotting, wake-word detection) on the “edge” using Intel’s Neural Compute Stick 2 and other Intel technologies.
Before joining IIIT-Bangalore, I was with Sonus Networks (Now Ribbon Communications) where we developed ‘Element Management’ systems for 4G-VOIP products. At Sonus, I developed knowledge of Linux systems, Perl, Python and Java.
I have a bachelor’s degree in Telecommunication Engineering from Visvesvaraya Technological University, Belgaum, India.
You can reach me at firstname.lastname@example.org