Image Classifier - Classify frontal faces into two classes: shows teeth and shows not teeth.
Technologies:
- Python 2.7
- Tensor Flow
- OpenCV
- Scikit-neuralnetworks
- Scikit-learn
- Scikit-images
- numpy
- Convolutional Neural Networks
- ImageNet
Training
python tf_trainer.py --bottleneck_dir=bottlenecks --how_many_training_steps=4000 --model_dir=inception --summaries_dir=training_summaries/basic --output_graph=retrained_graph.pb --output_labels=retrained_labels.txt --image_dir=teeth_images_org/ --learning_rate=0.005
Classification
python tf_classifier.py ./B ./results
Training the CNN
python sknn_trainer.py ./all_teeth_org/ cnn_graph.nn ov 0
Classification
python sknn_classifier.py ./B ./results cnn_graph.nn
./B
is the directory with the images to classify
./results
is the directory where the classified images will be saved, one sub-directory for each class