AI Glasses for Low‑Vision Users
A wearable system that combines TinyML object detection, depth estimation, and conversational assistance to help users navigate and recognize objects in real time.
Research engineer with experience across IC design, embedded ML, and product prototyping. I enjoy the place where algorithms, hardware, and human needs meet to tell useful stories.
A wearable system that combines TinyML object detection, depth estimation, and conversational assistance to help users navigate and recognize objects in real time.
End‑to‑end acceleration of a lightweight UNet with a Mamba bottleneck. Focus on quantization, memory traffic, and selective‑scan hardware for efficient inference.
Won multiple awards; co‑authored a paper on an EEG‑based interface for investigating Bach’s effect on Tourette’s syndrome patients (VLSI/DAT 2024).
RISC‑V Creativity Competition (3rd Prize, Application). A PINN approach for partial differential equations with custom training pipeline.