TensorflowSamples

classifier which is trained on Pepso Coke and Sprite bottles images .feed image to classify among 3 category .

Retrain based on parameters.

python -m scripts.retrain \
--bottleneck_dir=tf_bevrage_files/bottlenecks \
--how_many_training_steps=500 \
--model_dir=tf_bevrage_files/models/ \
--summaries_dir=tf_bevrage_files/training_summaries/"${ARCHITECTURE}" \
--output_graph=tf_bevrage_files/retrained_graph.pb \
--output_labels=tf_bevrage_files/retrained_labels.txt \
--architecture="${ARCHITECTURE}" \
--image_dir=tf_bevrage_files/bevrage_photos

Classification

python -m scripts.label_classification \
--graph=tf_bevrage_files/retrained_graph.pb  \
--image=tf_bevrage_files/bevrage_photos/sprite/093187.jpg