Repository files navigation

Can you virtually remove a face mask to see what a person looks like underneath? Our Machine Learning team proves it’s possible via an image inpainting-based ML solution.Here is exactly how our engineers approached the problem — from the preconditions to the implementation, results and future improvements.

Check outthe article for a more in-depth explanation

Examples of results (input | expected output | actual output)

Results can be replicated by following those steps:

• If yoursystem does not have an Nvidia CUDA device available, please commenttensorflow-gpu==2.2.0 in theenvironment.yml file.
• If you are running MacOS, changetensorflow==2.2.0 totensorflow==2.0.0 in theenvironment.yml file.
• UseConda ENV Manager to create new ENV:conda env create -f environment.yml
• Activate the ENV:conda activate mask2face

• DownloadLabeled Faces in the Wild data and unzip its content intodata folder or usemask2face.ipynb that will download it automatically.
• You can get better results using larger dataset or dataset with higher quality images. For exampleCelebA dataset contains over 200 000 high quality images.

You can configure the project using theconfiguration.json. Some of the items are set up and should not be changed. However, changing some of the following items can be useful.

• input_images_path: Path to dataset with images that are input for the DataGenerator. If you want to use a different dataset than the default one, set the path to it here.
• train_data_path: Path where are training images generated and where training algorithm is looking for training data.
• test_data_path: Path where are testing images generated.
• train_image_count: Number of training image pairs generated by DataGenerator.
• test_image_count: Number of testing image pairs generated by DataGenerator.
• train_data_limit: Number of training image pairs used for model training.
• test_data_limit: Number of testing image pairs used for model testing.

• Run Jupyter server in the ENV:jupyter notebook
• Follow instructions in themask2face notebook

• There might be a memory leak when training the model
• Performance on real-world data might not be optimal - follow ourtips to get possible results on real-world data
• If you encounter any other issues, please, let us know; we are happy to help

If you’re considering our help, you may be interested in our other past work—like thecustom AI solution we built for Cinnamon in just four months. And if you’re a fellow engineer, please feel free to reach out to us with any questions or share your work results. We’re always happy to start up a discussion.