- NCSDK2 installed to compile model and use its Python API.
- Keras installed for training the model.
pip3 install keras
- Optionally, copy this folder into your NCSDK2 directory along with other TensorFlow examples.
ncsdk/examples/tensorflow/keras_mnist - Plug NCS to a USB port on the host machine.
- Run command -
make all - Run command -
make run
Provided Makefile describes various targets that help with the above mentioned tasks.
Runs Keras model training(only for the first time), ncprofile, nccheck, nccompile.
Train and save Keras MNIST model if Keras model files are not found.
Runs the provided network on the NCS and generates per layer statistics that are helpful for understanding the performance of the network on the Neural Compute Stick.
Uses the network description and the trained weights files to generate a Movidius internal 'graph' format file. This file is later used for loading the network on to the Neural Compute Stick and executing the network.
Runs the provided predict-mnist-ncsdk2.py file which sends a single image to the Neural Compute Stick and receives and displays the inference results.
Runs the provided predict-mnist-ncsdk2-cam.py file which is similar to previous demo but take live webcam image as input.
Runs the network on Caffe on CPU and runs the network on the Neural Compute Stick. Check then compares the two results to make sure they are consistent with each other.
Removes all the temporary files and trained model files that are created by the Makefile
After training the Keras MNIST model, 3 files will be generated, while the conversion script convert-mnist.py only use the first 2 files to generate TensorFlow model files into TF_Model directory.
model.json Only contain model graph (Keras Format).
weights.h5 Only contain model weights (Keras Format).
model.h5 Both contain model graph and weights (Keras Format).