Build a TensorFlow Image Classifier in 5 minutes [2019]
Learn how to create your own image classifier in only 5 minutes! This video teaches you a quick and easy way to get started with TensorFlow. The method still works in 2019!
Here are all the steps that are covered in the video:
Step 1: Download TensorFlow
pip install tensorflow
Step 2: Clone Git repository
git clone https://github.com/googlecodelabs/ten...
Step 3: Download training images
curl http://download.tensorflow.org/exampl... \
| tar xz -C tf_files
Step 4: Define variables
IMAGE_SIZE=224
ARCHITECTURE="mobilenet_0.50_${IMAGE_SIZE}"
Step 5: Train the classifier
python -m scripts.retrain \
--bottleneck_dir=tf_files/bottlenecks \
--how_many_training_steps=500 \
--model_dir=tf_files/models/ \
--summaries_dir=tf_files/training_summaries/"${ARCHITECTURE}" \
--output_graph=tf_files/retrained_graph.pb \
--output_labels=tf_files/retrained_labels.txt \
--architecture="${ARCHITECTURE}" \
--image_dir=tf_files/XXX_photos
Step 6: Prediction
python -m scripts.label_image \
--graph=tf_files/retrained_graph.pb \
--image=tf_files/XXX.jpg
Let me know if you have any questions, and I will do my best to help you!
Subscribe to become a coder in minutes!
/ @coderinminutes9591