Developed and trained an AI model to recognize various types of waste on a resource-constrained GAP9 microcontroller from Greenwaves Technologies.
I used the WaRP-D dataset for training and optimized the model for real-time, on-device inference.
The system uses a camera as input for live detection, requiring extensive optimization to ensure efficient performance on the hardware.
| What I worked with | Details |
|---|---|
| Hardware | GAP9 microcontroller (Greenwaves Technologies), Camera |
| Dataset | WaRP-D |
| Technologies | Embedded AI, Model Optimization, Real-Time Inference |
| My Role | AI Development, Model Training, Embedded Deployment |
| Impact | Enabled real-time waste detection with limited computational resources |