A music visualizer which creates visualizations via stable diffusion XS and audio features via librosa.
A competition to predict plant traits from images and tabular data. My implementation uses image feature extractor and FiLM blocks...will make a blog post when im done.
Created a project with a Streamlit frontend that takes an English YouTube video and converts the English audio to German. Also provides German subtitles.
Published on: Fall 2023
Project that generates code for visualizations that the user describes or uploads a picture of. Allows for fine-tuning of the visualization with more prompting.
Published on: Spring 2023
Implementing dataset description generation through contextual prompting.
A project where experimentation was done to see if claims made on Graph Contrastive Learning are true, and if they can be improved.
Published on: Fall 2022
An augmented reality app that detects objects around the user (YOLOv3), generates activities, and gives them to the user to distract from nicotine cravings.
A blockchain implementation that adds a new block containing song information when a new song is played on Spotify.
Published on: Fall 2021
A philosophical text portal that aims to help users understand and connect their ideas using features supported by GPT-3.
Implementing and improving intersection tracking and counting on vehicle trajectories, using Deepsort + Yolov5. This project is a Town + Gown project collaborated with DOT and DDC
Curating a low fidelity image dataset for traffic object detection, ensuring all classes of objects are represented correctly.
Implemented solo RAFT consensus protocol for the ledger.
Published on: July 2021
Worked as the machine learning head for the nanodiamonds subteam of the NYU AI for Scientific Research Vertically Integrated Project.
Published on: June 2021
Implementing illegal parking detection on low-quality traffic cameras.
Published on: December 2020
Implementing social distancing violation detection on low-quality traffic cameras.
Published on: August 2020
Performed image preprocessing for OCR on low-quality and messy images.
Published on: June 2020
