yumatsuoka/ikemen_check

regression a score of good looking face with convolutional neural networks and face image datasets I made.

#IkemenCheck:Regression or Classification with CNN ##Abstract regression or classification with Convolutional Neural Networks using TensorFlow.

In 'src/cnn/.', "get_image_tensor.py" is to make a object of images dataset and classification.py and regression_cnn are CNNs codes.
In 'src/implement_to_datasets', ".sh" or ".py" are to transform images.

##Requirements Implement CNNs
-Python 2.x (checked Python 2.7.6)
-TensorFlow(checked TensorFlow 0.6.0)
-Pillow(checked Pillow 3.0.0)
-numpy(checked numpy 1.11.0)
-pandas(checked pandas 0.17.1)

Implement transforming images
-OpenCV(checked OpenCV 3.0)
-pandas(checked pandas 0.17.1)

##How to use -prepare datasets(.jpg) and target_data(.csv) with 'src/implement_to_datasets/.'scripts
-python classification.py or regression_cnn.py