Courses Detail Information
ECE4880J – Computer Vision
Instructors:
Credits:
4
Pre-requisites:
ECE2810J
Description:
This course introduces the foundations and modern practice of computer vision. Students learn core tasks—image classification, object detection, image segmentation, video understanding, and 3D vision—alongside the deep learning models that power them, including CNNs, RNNs, attention mechanisms, and Transformers. The class blends classical approaches with state-of-the-art techniques, with hands-on labs and a semester-long team project that emphasize practical implementation and evaluation using PyTorch.
Course Topics:
1. History and scope of computer vision
2. Image classification
3. Optimization and regularization
4. Recurrent models for visual sequences
5. Attention and Transformers
6. Object detection and image segmentation; model interpretation
7. Video understanding
8. Large-scale/distributed training
9. Self-supervised learning
10. Generative models
11. Vision–language methods
12. 3D vision
13. Robot learning
14. Medical imaging applications