Development of an improved generative adversarial network method for data augmentation and its application in environmental and financial domains

Using simplified language understandable to a layperson; provide a general, one-paragraph description of the proposed research project to be undertaken by the intern(s) as well as the expected benefit to the partner organization. (100 – 150 words) This project aims to increase image datasets by not doing experiments or collecting physical checks. Instead, the image data augmentation is implemented by generative adversarial networks (GANs), generating new images from original images using different algorithms. GANs have a generator and a discriminator. The generator generates new images from random inputs, and the discriminator calculates the loss between new images and original images. The weights of networks will be adjusted through back-propagation and to reach minimal loss between new and original images. The results can benefit the partner organization in augmentation check images to free the model’s performance from the limitation of datasets.

Faculty Supervisor:

Bing Chen

Student:

Yifu Chen

Partner:

Verafin Inc.

Discipline:

Engineering

Sector:

Information and cultural industries

University:

Memorial University of Newfoundland