Inspiration
I have heard multiple times of farmers facing problems with crops and this could be an awesome tool to share images and know which disease a plant has.
What it does
Uses Computer Vision to classify between 38 diseases from plant images.
How we built it
I used pre-trained weights to transfer learn on top of and created an upstream model with some CNN layers and used knowledge distillation to get better performance
Challenges we ran into
Prevent overfitting since only 3 GB data, 88000 images were available
What we learned
Learned about model pruning
Built With
- python
- tensorflow
Log in or sign up for Devpost to join the conversation.