Yiming Huang

Incoming Summer 2026

- Developed advanced Python scripts to streamline the analysis of datasets comprising over 1,000 cell images, providing crucial data to the study of Parkinson’s Disease - Leveraged computer vision techniques and Python libraries (OpenCV, scikit-image) to perform cell segmentation and quantification tasks - Co-authored multiple peer-reviewed research publications

- Composed and integrated 150 musical ear training exercises in the application for 1,000+ music learners - Engineered Python scripts to interpret and convert MusicXML files into other formats (png, mp3, Lilypond), allowing continuous generation of 500+ resource files - Built a Python library for analyzing musical information (pitch, interval, chord)

- Improved T-cell receptor binding prediction accuracy by 30% by training a TensorFlow neural network, contributing to advancements in cancer drug discovery and disease mechanism research - Fine-tuned Large Protein Language Models (ProtBert, ESM2) on a curated dataset of 50,000 TCR-epitope pairs to enhance model performance - Conducted comprehensive evaluations of existing models (NetTCR-2.0, VIBTCR), identifying key opportunities for performance optimization - Implemented a robust data-cleaning pipeline for the dataset, ensuring high-quality input for model training - Performed extensive data analysis based on metrics like accuracy, AUC, and F1 score, and visualized results using Matplotlib