ML course project, including data analysis part, implement network and final submission (with accuracy around 0.94).
Kaggle competition home page: https://www.kaggle.com/c/plant-pathology-2020-fgvc7
ML course project, including data analysis part, implement network and final submission (with accuracy around 0.94).
This is a really simple neural structure (use the pre-trained network with an additional dense layer to produce label output).
Friendly with CPU/GPU, you can train it in Mac (reach accuracy of 90% within 40 min).
If it doesn’t work for you due to the lack of C/GPU memory, reduce the batch size.