Emotion Detection from Tom and Jerry videos

This is an application to detect emotions from Tom and Jerry videos.

Installation

The repository can either be cloned or downloaded as a zip.

Usage

UPDATE: No more complex setup! Refer READMEDocker.md for instructions to easily set it up using Docker :)

The trained yolo detector and facial expression detector's weights should be first downloaded from the following links :

YOLO Detection module (1)

Facial Expression module (2)

The YOLO Detector weights (1) should be placed inside the yolo_predictor directory and the Facial Expression weights (2) should be placed in the root directory.

As long videos may take longer time, it would be better to convert the video into small clips as in

ffmpeg -i movie.mp4 -ss 00:00:03 -t 00:00:08 -async 1 cut.mp4

(refer : https://stackoverflow.com/questions/18444194/cutting-the-videos-based-on-start-and-end-time-using-ffmpeg)

To run on the default video cut.mp4, following command can be run from the root directory

python main.py

To run on some other video (ex : movie.mp4 )

python main.py -i movie.mp4

OpenCV allows us to generate videos with no audio. Hence the audio has to be extracted separately and combined with the output video file. We use ffmpeg for this purpose.

For generating audio file from the input video

ffmpeg -i filename.mp4 filename.mp3

For combining the audio and generated video

ffmpeg -i video.mp4 -i audio.mp3 -c:v copy -c:a aac output.mp4

Implementation

The details of implementation can be found in Implementation.md

Result

Click above picture to view video

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.

Please make sure to update tests as appropriate.

License

MIT

Thanks for reading :)