Any projects related to the use of deep learning that involves reasonable amount of efforts will be accepted. You are expected to submit a project report, and deliver a 10-15 minutes presentation (including answering questions from the class). The following are some example projects:
1) Train a deep learning model that can estimate the age of a given photo. The training set consists of a large number of human photos along with ages.
2) With the same training set, train a deep generative model that can generate an image of the person at certain age when given the image of the same person at certain age.
3) With a collection of images along with captions, train a deep neural network that can generate caption for new images.
4) Train a deep model that can tell if a certain document (e.g., novel) was written by Shakespeare, given a collection of work authenticated to be written by Shakespeare.
You can look up internet resources for related work. Group projects are highly encouraged (group size limited by 4). Project report will be graded by both quality of work and clarity of presentation.