ECE 285 MLIP: Machine learning for image processing

This class focuses on recent advances in machine learning (especially deep learning) and their applications to image processing and computer vision. After introducing some of the fundamentals of statistical learning (1½ week), we will discuss about neural networks and support vector machines and their optimization (2½ weeks). We will next move to deep learning and enumerate the numerous algorithmic recipes that led to its success (2 weeks). We will next focus on image classification, introduce Convolutional Neural Networks, and review successful architectures (1½ week). Applications to object detection, image segmentation, image captioning, image generation, super-resolution and style transfer will finally be discussed (2½ weeks).

There will be 4 assignments, 1 project and 3 quizzes.