This project is mainly based on VGG16 and Inception V3.
Motivation comes from life. When describing a person, often several features will be listed, for example, he/she has arched eyebrows and brown hair, or wears wavy hear and has high cheekbones. So we are thinking about if a mechanism that can also generate such features which are usually described by a human. Training deep learning models require large, hard-to-come datasets to perform the task to achieve optimal results. So here we load the pre-trained well-established network and train it on the CelebFaces Attributes Dataset (CelebA). We aim to build a network which take a photo as input and generate the facial feature of that person.
https://drive.google.com/drive/folders/1s4Wi4t7MjpmZtS_gx01pLTYMZopJoWGM?usp=sharing
https://drive.google.com/drive/folders/1ZEBrZTZql4FfzsPs026NWtf1khw4czrL?usp=sharing