Name: Shanti Stewart
Type: User
Company: University of Southern California
Bio: Recent M.S. (ML) grad @ USC.
Current research intern @HumeAI.
Research interests: audio & music generative modeling, text-to-speech, multimodal learning.
Location: New York City
Shanti Stewart's Projects
Community list of startups working with AI in audio and music technology
deepinterpolation is a Python library to denoise data by removing independent noise. Importantly training does NOT require ground truth. This repository is currently meant to support the bioRxiv publication results : https://www.biorxiv.org/content/10.1101/2020.10.15.341602v1
Research project on hand movement decoding from EEG signals using deep learning and signal processing algorithms.
Emo-CLIM: Emotion-Aligned Contrastive Learning Between Images and Music [ICASSP 2024]
A class research project on independent component analysis (ICA) and FastICA for Matrix Analysis for Signal Processing and Machine Learning (ECE 599) course at Oregon State University.
This repository contains the implementations of the MUSIC and ESPRIT algorithms, which can be used for super-resolution spectral analysis.
This github repo is for Neurips 2021 and Interspeech 2022 papers on Non-Matching Reference based estimation of speech quality assessment.
Computer vision algorithms implemented in C for a project involving obstacle avoidance for small mobile robots.
A configurable version of the SampleCNN model ("Sample-level Deep Convolutional Neural Networks for Music Auto-tagging Using Raw Waveforms"), implemented in PyTorch.
Repository with a profile README file.
Final project for Deep Learning 1 (EE 541) course at USC.
Final project for Machine Learning 1 (EE 559) course at USC.