Deep Learning for Computer Vision

CMPE 593

Instructor: Asst. Prof. Berk Gökberk

Credits: 3

Course Description

Graduate course on applied deep learning for computer vision: convolutional networks, detection, segmentation, generative models, and vision transformers.

Course description

This course offers an applied approach to teaching modern deep learning–based computer vision concepts. Students learn and use contemporary techniques to solve fundamental vision problems. Deep neural network architectures are covered for tasks such as object detection, recognition, and segmentation.

Learning outcomes

After successful completion of this course, students will be able to:

  1. Use convolutional neural networks.
  2. Apply deep learning–based computer vision methods to solve problems.
  3. Evaluate the performance of computer vision models.

Topics (tentative)

  • Introduction to deep neural networks for computer vision
  • Convolutional neural networks
  • Deep image models
  • Object detection using DNNs
  • Image segmentation using DNNs
  • Deep generative vision models
  • Temporal vision models: vision transformers

Course materials

Course content and announcements are managed through Moodle; university e-mail is used for communication. Contact the instructor by e-mail to schedule office hours (CMPE Building, 2nd floor, room 23).