------- Written by Daniel Ogunlolu -------- INFO: Date: 30, NOV 2021 Programming Language: Python 3 Main Script: extract-face.Python Docker Base Image: orgoro/dlib-opencv-python ALGORITHM: 1. Program starts by Downloading the given video from youtube link: https://www.youtube.com/watch?v=JriaiYZZhbY 2. Saves video inside directory (data/) with extension .mp4 3. Run HOG face detection on each frame, all detections are saved in data/faces directory 4. Finally, a Clustering process is performed on all faces to filter unique faces 5. Unique faces are saved in data/unique_faces directory 6. When Program is done, enter q to end and exit the dockeer container NOTE: After building and running the docker image/container, you can copy directory containing faces extracted (data) using the following command: docker cp <container id>:/main/data/ <host/path/to copy to/> To view container id, use the following command: docker ps replace <host/path/to copy to/> with path where you want the data directory to be copied to. eg: docker cp ad9f3d69bde2:/main/data/ /home/amazing/Desktop RESULT: Tested with the given video; the program did not work so well due to the fact that the given video has a bad quality and most of the faces appeared blurry (and very small). Tested the program with a diffrent video which has a better quality (faces were more vivid and clear) and the program worked well. I have attached a zip files containing the final result (unique-faces) ---------------- DOCKER HOW TO: