Fnu Sidharth Email & Phone Number

Ann Arbor, MI, US

Laguna Hills, California, United States

- Working on state space models (TENNS 2.0) for speech enhancement in noisy environments and ASR- Worked on aTENNuate: Deep state space model for real-time speech enhancement, submitted to IEEE ICASSP 2025

Seattle, Washington, United States

Neural Systems Lab, GridLab, Herron LabWorking with Dr. Jeff Herron, Dr. Rajesh Rao, Vishwas Sathish, and Dr. Samantha H Sun on finding potential biomarkers of pain from electrophysiological signals using data driven approaches.

Seattle, Washington, United States

• Advisors: Dr. Shyam GollakotaSummer 2024Project: Adaptive enrollment for Target Speech Extraction
• Excited to be working on extending the "Look Once to Hear: Target Speech Hearing with Noisy Examples" paper! I'm developing model to extract conversation mixture with volume controlled target speaker speech. This.

Seattle, Washington, United States

• Advisors: Dr. Shyam Gollakota, Malek ItaniSpring 2024 Project: Conformer-Enhanced Acoustic Source Localization and Masked Channel Prediction in Noisy Environments
• Investigated sound prediction using a 4-microphone array in a fixed square position with a random sound source in a room
• Developed a self-supervised model based on a conformer architecture. Modeled room acoustics using gpuRIR and created a pipeline from data creation to masked channel prediction.
• Achieved a validation mean squared error (MSE) of 0.001 in reconstructing the masked microphone signal.

Bengaluru, Karnataka, India

• Developed and evaluated speaker and language diarization systems for multilingual, multi-speaker environments with code-mixing
• Collaborated with a team of ten to pre-process conversational audio signals using Audacity and Praat, fine tune Speech Activity Detector (SAD)based on x-vectors, and cluster the speakers using AHC and SC with VB-HMM.
• Achieved a Diarization Error Rate (DER) of 28.04 for speaker diarization and 37.72 for language diarization baselines on the DISPLACE Challengedataset, showcasing the effectiveness of the proposed algorithm

Trivandrum, Kerala, India

— Worked on "Emotion detection from EEG using transfer learning" — Supervised by prof Dr. Jerrin Thomas Panachakel— Presented the work at IEEE EMBC 2023, IEEE MetroXRAINE 2023

Trivandrum, Kerala, India

— Worked on Multilabel Bird Call Classification using Graph Convolutional Networks — Guided by prof Dr. Rajeev Rajan

Master Of Science - Ms, Electrical And Computer Engineering, 3.9

Relevant courseworks -Computer Vision -Introduction to statistical learning -Computer Speech Processing -Digital Signal Processing

Bachelor Of Technology - Btech, Applied Electronics And Instrumentation, Mathematics, 9.35

Activities and Societies: Created a first ever mathematical society in Campus : MATHLETES