Contributors
- Weijie Lyu
- Sirui Wang
The final project for course CS-445 Computational Photography at the University of Illinois, Urbana-Campaign.
In this project, we detected emotions of human faces and replaced the faces with corresponding emojis. For the faces whose orientations are not forward, we detected the orientations of the faces and applied transformations to emoji images. So that the orientations the emoji images can be the same as the faces.
The images you want to process can be put in img folder. To run the experiment, simply run the face2emoji.ipynb
We used AffectNet dataset from Kaggle to train our emotion classifier.
[1] https://github.com/timesler/facenet-pytorch
[2] Mollahosseini, Ali, Behzad Hasani, and Mohammad H. Mahoor. "Affectnet: A database for facial expression, valence, and arousal computing in the wild." IEEE Transactions on Affective Computing 10.1 (2017): 18-31.
[3] https://www.kaggle.com/datasets/tom99763/affectnethq
[4] https://pyimagesearch.com/2017/04/03/facial-landmarks-dlib-opencv-python/