Glimpsees is Python application that tracks and extracts faces from videos, saving them as images. You can also compile these extracted faces into an animated GIF.
Glimpsees is designed for Google Colab, but you can run it on your local machine, just comment out the following lines:
- App section: Lines 5 and 108
- GIF section: Line 37
-
Upload Your Video File: Begin by uploading your video (MP4 and AVI formats are currently supported) to your project directory.
-
Install Libraries: Run the Libs command to ensure all necessary dependencies are installed.
-
Configure File Name: In the Pre-config section, update the file_name variable to match the exact filename of your uploaded video (e.g.,
my_example_video.mp4). This name will be used for your output files (e.g.,faces_my_example_video.zip,faces_my_example_video.gif). -
Set Custom FPS (Optional): If you wish to extract faces at a different frame rate than the original video, you can set a custom value for FPS. Leave this setting unchanged to use the video's original frames per second.
-
Run the Application: Execute the App script. Glimpsees will then process your video, detect first face (you can change it on line 65), and save each detected face per frame as a separate image.
-
Generate GIF (Optional): After the App script completes, you can run the GIF script to compile all the extracted face images into an animated GIF.
Glimpsees is an open-source project. Feel free to fork the repository, explore the code, and contribute to its development!
This project is licensed under the MIT License – see the LICENSE file for details.