See our latest COVID updates.

GANs for Data Synthesis | AISV.809

GANs for Data Synthesis | AISV.809


Technology Summary

GANs (Generative adversarial networks ), a subset of generative models, are one of the most helpful deep learning techniques in recent years, particularly for the tasks of data synthesis. GANs allow us to estimate an intrinsic data distribution from a given dataset and generate newer data that look like one from the given dataset.

Who should take this course

This is a course for deep learning students interested in advanced computer vision, image synthesis, denoising, super-resolution, language modeling and text generation, anomaly detection, drug molecules synthesis, deep reinforcement learning, control systems, and autonomous driving.

Course Description

In Generative Adversarial Networks for Data Synthesis, students explore the theoretical and mathematical framework of GANs along with hands-on guided workshops and practical applications in image synthesis space. They learn the fundamental concepts of generative models, mathematical formulation, and practical aspects of building and training them using TensorFlow and Keras.

Working in a research environment, you’ll learn the problems and challenges associated with GANs and overcome them at the production level. You will implement image synthesis from image and text and drug molecule synthesis using state-of-the-art GAN-based DNN architectures You’ll implement deep learning algorithms from technical papers for deep generative models and focus on building an intuition of efficient training of DL and GAN models.

Learning Outcomes
At the conclusion of the course, you should be able to:

  • Define the generative models and their use cases
  • Build the generative adversarial networks (GANs) framework
  • Train GANs with variants of their cost functions
  • Explain the generator, discriminator, GAN loss function, and adversarial loss
  • Build conditional GANs and WGAN
  • Perform image translation and synthesis tasks with state-of-the-art networks, such as Pix2Pix, CycleGAN

Time commitment
  • Total instruction: 30 hours
  • Weekly estimate: 3 hours lecture and 6–8 hours out-of-class study

Requisite Knowledge
Skills needed: Familiarity with probability theory and linear algebra, programming, deep learning Recommended course: Deep Learning and Artificial Intelligence with TensorFlow and Keras (DBDA.X425)

Have a question about this course?
Speak to a student services representative.
Call (408) 861-3860
FAQ
ENROLL EARLY!
This course is related to the following programs:

Prerequisite(s):

Sections Open for Enrollment:

Open Sections and Schedule
Start / End Date Quarter Units Cost Instructor
02-07-2023 to 04-11-2023 3.0 CEUs $980

Ajay K Baranwal

Enroll

Final Date To Enroll: 02-07-2023

Schedule

Date: Start Time: End Time: Meeting Type: Location:
Tue, 02-07-2023 6:30 p.m. 9:30 p.m. Flexible SANTA CLARA / REMOTE
Tue, 02-14-2023 6:30 p.m. 9:30 p.m. Flexible SANTA CLARA / REMOTE
Tue, 02-21-2023 6:30 p.m. 9:30 p.m. Flexible SANTA CLARA / REMOTE
Tue, 02-28-2023 6:30 p.m. 9:30 p.m. Flexible SANTA CLARA / REMOTE
Tue, 03-07-2023 6:30 p.m. 9:30 p.m. Flexible SANTA CLARA / REMOTE
Tue, 03-14-2023 6:30 p.m. 9:30 p.m. Flexible SANTA CLARA / REMOTE
Tue, 03-21-2023 6:30 p.m. 9:30 p.m. Flexible SANTA CLARA / REMOTE
Tue, 03-28-2023 6:30 p.m. 9:30 p.m. Flexible SANTA CLARA / REMOTE
Tue, 04-04-2023 6:30 p.m. 9:30 p.m. Flexible SANTA CLARA / REMOTE
Tue, 04-11-2023 6:30 p.m. 9:30 p.m. Flexible SANTA CLARA / REMOTE