Final Project for Computer Vision course AY2024 Spring

This project aims to convert video files into transparent background GIFs using 3D Markov Random Field. It provides an easy-to-use GUI interface which allows the user to select his/her region of interest to segment. (similar to GrabCut)
- Clone the Github Repository and
cdinto the directorygit clone https://github.com/minsuk00/VideoToGIF cd VideoToGIF - Install the required dependencies using package manager (activate virtual env if necessary)
pip install -r requirements.txt
- Run application
python main.py
- After loading video (mp4 file), move the frame to the desired start point
- Draw a rectangle, and draw lines to indicate foreground and background

- Process image to construct a GMM for foreground and background. (Iterate 5-6 till result is satisfactory)

- Run video segmentation & download as GIF (below is an example result case)




