M.S. in Computer Science @ UC San Diego · Class of 2027 · Machine Learning Systems · AI Agents · Full-Stack Products
I’m Yuchen (Eugene) Song, an M.S. Computer Science student at UC San Diego (Class of 2027). I build machine-learning systems, AI agents, and full-stack products, with research in wearable sensing, audio modeling, and medical computer vision.
- Research: wearable accelerometer modeling, audio ML, and medical-image segmentation
- Engineering: Python, PyTorch, TypeScript, React Native, Next.js, Firebase, and cloud services
- Currently: contributing to CHAP2.0, a reproducible posture-classification pipeline for wearable data
| Project | What I worked on |
|---|---|
| DeepPostures / CHAP2.0 | PyTorch workflows for zero-shot prediction and fine-tuning of hip- and wrist-worn accelerometer data. |
| MyGradCard | Cross-platform fundraising product built with React Native, Next.js, Firebase, and Stripe. |
| MADUV 2025 | Interspeech 2025 implementation for mouse ultrasonic-vocalization classification with pretrained audio models. |
| Laryngoscopic Image Segmentation Toolkit | Co-author of a SAM prompt-engineering toolkit for vocal-fold and glottis segmentation. |
- Y. Song, Y. Zhang, M. Li. Exploring Pre-trained Models on Ultrasound Modeling for Mice Autism Detection with Uniform Filter Bank and Attentive Scoring. Interspeech 2025.
- Y. Zhang, Y. Song, J. Liu, M. Li. An Automatic Laryngoscopic Image Segmentation System Based on SAM Prompt Engineering: From Glottis Annotation to Vocal Fold Segmentation. Frontiers in Molecular Biosciences, 2025.
Open to research and engineering collaborations.