I build machine-learning systems for human movement, clinical sensing, and embodied AI. My work connects computer vision, EEG, motion capture, wearable sensors, and robotics with a practical goal: measurement tools that clinicians and researchers can trust.
Focus
Vision-language coaching
Fine-tuning multimodal models to assess ingestive behavior and generate clinician-style feedback from meal video.
Neural and motion decoding
Synchronizing high-density EEG with motion capture for movement decoding and BCI-oriented analysis.
Clinical movement assessment
Building marker-free computer-vision and sensor pipelines for rehabilitation and neuromuscular measurement.
Current Work
Vision-language models for ingestive-behavior coaching
Fine-tuning multimodal models to jointly classify eating behavior and generate clinician-style coaching feedback.
Electrode media comparison and VEP analysis
Building reproducible EEG analysis pipelines for resting-state, alpha-reactivity, and visual-evoked-potential recordings.
Perception for LLM-based human-robot teaming
Developing computer-vision perception modules for closed-loop cobot autonomy allocation.
Synchronized movement decoding
Aligning high-density EEG with optical motion capture for grasping and reaching analysis.
Toolkit
PyTorch, Hugging Face, LoRA fine-tuning, scikit-learn
OpenCV, YOLO, segmentation, pose estimation, video analysis
MNE, EEG preprocessing, EMG analysis, spectral features
Python pipelines, data synchronization, experiment tooling
Contact
For research collaborations, project questions, or shared interests in ML for clinical sensing, email is the best way to reach me.