Image Augmentation for 3D Images

This is a simple example of image augmentation for 3D images using Analytics ZOO API. We use various ways to transform images to augment the dataset.

Environment

• Python 3.5/3.6 (numpy 1.11.1)
• Apache Spark 2.x (This version needs to be same with the version you use to build Analytics Zoo)

Install or download Analytics Zoo

Follow the instructions here to install analytics-zoo via pip or download the prebuilt package.

Run after pip install

You can easily use the following commands to run this example:

export SPARK_DRIVER_MEMORY=1g
jupyter notebook --notebook-dir=./ --ip=* --no-browser

See here for more running guidance after pip install.

Run with prebuilt package

Run the following command for Spark local mode (MASTER=local[*]) or cluster mode:

export SPARK_HOME=the root directory of Spark
export ANALYTICS_ZOO_HOME=the dist directory under the Analytics Zoo project

• Run the following bash command to start the jupyter notebook. Change parameter settings as you need, ie MASTER = local[physcial_core_number].

MASTER=local[*]
${ANALYTICS_ZOO_HOME}/bin/jupyter-with-zoo.sh \
    --master ${MASTER} \
    --driver-memory 1g  \
    --executor-memory 1g

See here for more running guidance without pip install.