Using Transfer Learning with VGG16, Resnet50, Inception v3 and Flask Framework to build a website that distinguishes 9 types of insects : armyworm, beetle, cicadellidae, cricket, grasshopper, limacodidae, lycorma delicatula, mosquito, weevil
Clone my project
git clone https://github.com/thangnh1/PestClassification
Open with editor tool, install lib with terminal
pip install -r requirements.txt
Train model (Recommend using Google Colab)
Upload folder PestClassification/data to Google Drive, then Open 3 file from PestClassification/notebook/ with Google Colab, connect GPU and Run All
Copy le.pkl to PestClassification and create folder models, copy 2 file .hdf5
from Google Drive to PestClassification/models/ . You can test models with predict.py
file or evaluate models with score.py
Open Terminal, run command python server.py