Responsibilities include:
- Conducted research on muscle fatigue characterization using surface electromyography (sEMG) signals for musculoskeletal disorder prevention.
- Designed and implemented an end-to-end preprocessing pipeline (filtering, MVC normalization padding sliding windows) for large-scale sEMG datasets.
- Adapted and extended Time-series Masked Autoencoder to extract high-level latent representations from noisy biomedical signals.
- Combined deep learning feature representations with handcrafted frequency-domain features to perform short window sEMG forecasting.
- Developed a modular and reusable codebase with MLflow integration for experiment tracking versioning and reproducibility.
- Collaborated with 3W Well With Waves on ergonomic applications providing AI driven insights for adaptive lead apron design to reduce muscle fatigue in healthcare professionals.