Skip to content

Deep convolutional generative adversarial network (DCGAN) trained to remove Snapchat filters from selfie images

License

Notifications You must be signed in to change notification settings

ipsingh06/ml-desnapify

Repository files navigation

Desnapify

Desnapify is a deep convolutional generative adversarial network (DCGAN) trained to remove Snapchat filters from selfie images. It is based on the excellent pix2pix project by Isola et al., and specifically the Keras implementation by Thibault de Boissiere.

The following figure shows a few examples of the output of the Desnapify on doggy filter selfies. Top row is the input, middle row is the Desnapified output, bottom row is the target.

Desnapification of the doggy filter

Desnapify currently includes a model trained to remove the doggy filter from images of size 512x512. It can be easily trained to remove other filters as well. This is left for future work.

Requirements

  • Python 3.5
  • Tensorflow (tensorflow-gpu with cuDNN is recommended)
  • Keras (Note: as of this writing the latest release of Keras 2.2.4 has a bug which prevents the model from being loaded. You must install the latest Keras from source as described here)
  • Python modules listed in requirements.txt

Setup

  1. Set up a virtualenv

    mkvirtualenv -p python3.5 desnapify
  2. Clone this repo

    git clone https://github.com/ipsingh06/ml-desnapify.git
  3. Install the python modules

    cd ml-desnapify
    pip install -r requirements.txt
  4. Install Keras from source as described here

Using the provided model to Desnapify images

This section describes how to use the provided trained model to remove the doggy filter from your images. The provided script src/models/predict_model.py takes care of scaling the images. We will be using the samples provided in data/raw/samples/doggy.

  1. Download the pre-trained model weights.

    python src/models/download_weights.py
  2. Run the prediction script. This will cycle through and display the resulting output images alongside their inputs.

    python src/models/predict_model.py \
        models/doggy-512x512-v1/gen_weights_epoch030.h5 \
        data/raw/samples/doggy \
        --image_size 512 512

    To store the results into images files, specify the --output <path to directory> option. You can also specify --no_concat to store the output image only.

    python src/models/predict_model.py \
        models/doggy-512x512-v1/gen_weights_epoch030.h5 \
        data/raw/samples/doggy \
        --image_size 512 512 \
        --output out \
        --no_concat

    Check the out directory for output images.

Training the model

This section describes how to train the model yourself. This requires generating pairs of images with and without the filter applied. The provided data generation script includes functionality to apply the doggy filter to images. It can be extended to apply other Snapchat filters as well.

The following figure shows the pipeline for generating training data from selfie images. Data pipeline

The provided model was trained with the People's Republic of Tinder dataset available on Kaggle. You should be able to use any dataset of selfie images for training.

Note: the training script uses multiprocess queues for batch loading, so you don't have to worry about fitting your entire dataset into memory. Use as large a dataset as you like.

  1. Place the set of selfie images in data/raw/dataset directory.

  2. Run the data generation script to generate image-pairs with and without the doggy filter.

    python src/data/make_dataset.py apply-filter \
        data/raw/dataset \
        data/interim/dataset \
        --output_size 512 512 \
        --no_preserve_dir

    This script will create two directories data/interim/dataset/orig and data/interim/dataset/transformed. The first contains images to be used as input to the model, and the second images to be used as the target.

  3. Run the data generation script again to split the dataset into training, validation and testing sets. This also packages all three datasets into a single HDF5 file.

    python src/data/make_dataset.py create-hdf5 \
        data/interim/dataset \
        data/processed/dataset.h5

    This script will create the file data/processed/dataset.h5

  4. We should verify that the dataset was created properly.

    python src/data/make_dataset.py check-hdf5 data/processed/dataset.h5
  5. Finally, we can run the training script.

    python src/models/train_model.py \
        data/processed/dataset.h5 \
        --batch_size 4 \
        --patch_size 128 128 \
        --epochs 30

    We can see how well the model is learning and performing by looking at current_batch_training.png and current_batch_validation.png images in reports/logs directory.

    We can also visualize the performance metrics using Tensorboard:

    tensorboard --logdir=reports/logs

    Tensorboard

  6. After training completes the model weights are saved in models/. To test the model, see the Using the provided model section.

Acknowledgments

  1. pix2pix project by Isola et al.

  2. Keras implementation of pix2pix by Thibault de Boissiere

  3. The apply-filters script is based off snapchat-filters-opencv

About

Deep convolutional generative adversarial network (DCGAN) trained to remove Snapchat filters from selfie images

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published