Scikit-image: https://scikit-image.org/
  Tensorflow: https://www.tensorflow.org/
  TCAV: https://github.com/tensorflow/tcav

python3 run_ace.py --num_parallel_runs 0 --target_class Zebra --source_dir SOURCE_DIR ---working_dir SAVE_DIR --model_to_run InceptionV3 --model_path PATH_TO_MODEL_CHECKPOINT --bottlenecks mixed_8 --num_test 20 --num_random_exp 40 --max_imgs 50 --min_imgs 30 --test_dir TEST_DIR 

num_parallel_runs: Number of parallel jobs (loading images, etc). If 0, parallel processing is deactivated.

target_class: Name of the class to be explained.

SOURCE_DIR: Directory where the discovery images (refer to the paper) are saved. 
It should contain (at least) two folders: 
1-"target_class" which contains images of the class to be explained. 
2-"random_discovery" which contains randomly selected images of the same dataset.

SAVE_DIR: Where the experiment results (both text report and the discovered concept examples) are saved.

model_to_run: One of InceptionV3 or GoogleNet is supported. You can change the "make_model" function in ace_helpers.py to have your own customized model.
model_path: Path to the model's saved graph.

TEST_DIR: Used for the profile classifier experiment (not part of the paper).
If None, the profile classifier experiment is not performed.
Same as source_dir:
1-"Name of the target class (here zebra)" which contains test images of the class to be explained. 
2-"random_test" which contains test images randomly selected from the test data.