This notebook classifies human facial emotions into 7 classes using fastai, and PyTorch. fastai is a layered API for deep learning; for more information, see the fastai paper and the vgg19_bn pre-trained model.
Refer to branches for comparison with Resnet18 and vgg16_bn Pre-trained models.
The code in the notebook is covered by the GPL v2 license; see the LICENSE file for details.
NB: fastai v1 currently supports Linux only, and requires PyTorch v1 and Python 3.6 or later.
Windows support is at an experimental stage: it should work fine but it's much slower and less well tested.
Dataset used in this notebook is based on fer_2013 and can be found here