Deployed a TensorFlow Lite BirdNet model on a Google Coral Board with an Edge TPU, building a solar-powered edge device that autonomously identifies bird species in the wild. Deployed classroom demonstrations on a NVIDIA Jensen Board.
Integrated the device with a citizen science bird database, enabling real-time wildlife sightings to be logged and contributed to broader ecological research.
Device
Hardware: Google Coral Dev Board (Edge TPU), NVIDIA Jensen Board
Model: BirdNet (TensorFlow Lite)
Power: Solar-powered edge device
Integration: Citizen science bird database for real-time wildlife sighting logs