Classify gender based on face image.
py -3 -m scripts.retrain --bottleneck_dir=tf_files/bottlenecks --model_dir=tf_files/models/"${ARCHITECTURE}" --summaries_dir=tf_files/training_summaries/"${ARCHITECTURE}" --output_graph=tf_files/retrained_graph.pb --output_labels=tf_files/retrained_labels.txt --image_dir=tf_files/gender
py -3 -m scripts.label_image --image=tf_files/gender/female/20_1_0_20170109213411083.jpg.chip.jpg
IMAGE_SIZE=299
tflite_convert \
--graph_def_file=tf_files/retrained_graph.pb \
--output_file=tf_files/optimized_graph.lite \
--input_format=TENSORFLOW_GRAPHDEF \
--output_format=TFLITE \
--input_shape=1,${IMAGE_SIZE},${IMAGE_SIZE},3 \
--input_array=Mul \
--output_array=final_result \
--inference_type=FLOAT \
--input_data_type=FLOAT
https://medium.com/datadriveninvestor/building-an-image-classifier-using-tensorflow-3ac9ccc92e7c
https://github.com/googlecodelabs/tensorflow-for-poets-2
https://codelabs.developers.google.com/codelabs/tensorflow-for-poets/#3
https://codelabs.developers.google.com/codelabs/tensorflow-for-poets-2-tflite/#2