Introduction

The course is designed for students who already have basic knowledge of neural networks and are proficient in Python (30 hours). It offers both theoretical and practical insights into deep learning, starting from foundational training methods to advanced applications in computer vision.

  • Luis Baumela, 16h x 2 Master = 30h
  • Roberto Valle, 16h x 2 Master = 30h

Unit 1: Introduction to Deep Neural Networks

This unit covers the fundamental techniques and concepts of training deep learning models:

  • Backpropagation algorithm.
  • Optimization algorithms such as SGD, Adam, RMSProp.
  • Regularization techniques including dropout, early stopping, and L2/L1 regularization.

The main goal is to provide a solid understanding of the training process and how to improve model performance and generalization.

Unit 2: Deep Learning for Computer Vision

This unit focuses on how deep learning is applied to computer vision problems:

  • Foundations of computer vision.
  • Convolutional Neural Networks (CNNs).
  • Popular architectures like AlexNet, VGG, ResNet.
  • Representation learning.
  • Applications: image classification, object detection, image segmentation.

It emphasizes practical use of deep learning to solve visual recognition and analysis tasks.