Speach to Sign Language

This repo uses the WLASL dataset along with google's mediapipe repo to train a model that can convert speach to sign language.

To Train

Download Videos

You can do this part on a desktop, but for speed's sake I used an IBM Virtual Server with 8 vCPUs.

• Create Virtual Server
• Follow the instructions here to install bazel
• Install opencv:

$ apt update
$ sudo apt-get install libopencv-core-dev libopencv-highgui-dev \
                       libopencv-calib3d-dev libopencv-features2d-dev \
                       libopencv-imgproc-dev libopencv-video-dev

• Install pip: sudo apt-get install python3-pip
• Install opencv-python: pip3 install python3-opencv
• Install youtube-dl
• Install ffmpeg: apt install ffmpeg
• Depending on system make sure python symlinks to python3 (i.e. Ubuntu): sudo ln -s /usr/bin/python3 /usr/local/bin/python
• Clone the repo: git clone https://github.com/RLashofRegas/speach-to-sign-language.git
• Run the download script. This will download and preprocess the videos using the WLASL repo:

$ cd speach-to-sign-language/
$ bash scripts/download_videos.sh

• Install tensorflow pip3 install tensorflow