Shrinidhi Mahesh
Masters of Science - Electrical Engineering (Machine Learning and Data Science)
University of Southern California
Hi there ! My name is Shrinidhi Mahesh and I am wrapping up a MS in EE from USC with my courses focused towards MLDS. My technical background combines Cloud Data Engineering production experience at Deloitte along with rigorous graduate coursework at USC in Deep Learning, Machine Learning, Reinforcement Learning and LLMs and Speech Processing. I have also been passionately training in Carnatic Music as a vocalist for the past 10+ years which drives my interest for Music Information Retrieval (MIR) and Audio ML.
Education
University of Southern California
Masters of Science in Electrical Engineering (Machine Learning and Data Science)
SRM Institute of Science and Technology
B.Tech. in Computer Science and Engineering
Publications
Experience
Research Assistant — University of Southern California
Analyst - Data Engineering — Deloitte Consulting
Portfolio
Foundation Model for Marine Mammal Vocalization and Classification [In Progress]
This project involved training a foundation model on a large dataset of marine mammal vocalizations, including whales and dolphins. The model was designed to learn general representations of the vocalizations, which could then be fine-tuned for specific tasks such as species classification and behavior analysis.
Emotion Recognition on MTG-Jamendo Dataset
A short experiment on classifying emotions in music using the MTG-Jamendo dataset.I trained a simple CNN model on the mel-spectrograms of the audio clips and achieved an accuracy of around 70%.The project is still a work in progress and I plan to explore more advanced architectures and techniques to improve the performance.
Collection of Carnatic Music Performances and recordings
Over the years, I have been actively involved in performing Carnatic music.This collection includes audio recordings of practice sessions of traditional compositions.
Collection of cover songs
A lot of random musings and bathroom singing sessions recorded over the years. This collection includes a mix of popular songs in various Indian languages.