This is an repository for 10701 machine learning course project
put extract_image_feature.py in the same directory of categories folders
create a category.txt with categories which you want to be pretrained in it
and split them by line break.
command:
- python extract_image_feature.py --model vgg16
- python read_features.py --model vgg16
Files | Descriptions |
---|---|
classify_image.py | classify images |
cnn_finetune_vgg16.py | fine-tune network |
extract_image_feature.py | extract features using vgg16 and so on |
load_data.py | load data |
read_features.py | read features from txt and train svm |
resnet50_best.h5 | weights after fine-tuning |
result.xlsx | result |
run.sh | for testing |