EECE 480EDeep Learning for Comp Vision

Catalog

This course explores modern deep learning approaches to computer vision, building on classical foundations to cover state-of-the-art neural architectures and methods. Topics include convolutional neural networks (CNNs) and modern CNN architectures, object detection and localization, semantic and instance segmentation, Vision Transformers, self-supervised and contrastive learning, multimodal vision-language models, generative models for vision, and video understanding. Students will implement and train deep vision models using PyTorch, and apply them to real-world tasks in autonomous systems, medical image analysis, and scene understanding. 3 Credits. Prerequisites: EECE 212 or equivalent with permission from instructor.