Generative AI

Hereby the curriculum per day:

• Day 1: Introduction to Generative AI

  Morning Session: Overview of AI and Machine Learning

  • Definitions and key concepts
  • Historical context and evolution of AI
  • Types of AI: Narrow vs. General AI

  Afternoon Session: What is Generative AI?

  • Definition and significance
  • Applications in various domains (art, music, text, etc.)
  • Ethical considerations and societal impact

• Day 2: Foundations of Generative Models

  Morning Session: Key Concepts in Machine Learning

  • Supervised vs. unsupervised learning
  • Basics of neural networks

  Afternoon Session: Types of Generative Models

  • Overview of generative vs. discriminative models
  • Introduction to Variational Autoencoders (VAEs) and Generative Adversarial Networks (GANs)

• Day 3: Deep Dive into GANs

  Morning Session: Understanding GAN Architecture

  • Generator and discriminator roles
  • Loss functions and optimization

  Afternoon Session: Training GANsChallenges and techniques (mode collapse, stability)

  • Hands-on: Building a simple GAN using TensorFlow/PyTorch

• Day 4: Exploring VAEs

  Morning Session: VAE Architecture and Components

  • Encoder-decoder structure
  • Loss functions: reconstruction loss vs. KL divergence

  Afternoon Session: Training and Applications of VAEs

  • Use cases (image generation, anomaly detection)
  • Hands-on: Implementing a VAE using TensorFlow/PyTorch

• Day 5: Advanced Generative Techniques

  Morning Session: Other Generative Approaches

  • Flow-based models and diffusion models
  • Introduction to Transformer-based generative models (e.g., GPT)

  Afternoon Session: Text Generation and Natural Language Processing

  • Overview of language models
  • Hands-on: Fine-tuning a pre-trained language model

• Day 6: Creative Applications of Generative AI

  Morning Session: Generative AI in Art and Music

  • Techniques for generating visual art and music
  • Case studies of notable generative projects

  Afternoon Session: Ethical Considerations in Creative AI

  • Copyright issues and ownership
  • Addressing bias and fairness

• Day 7: Evaluating Generative Models

  Morning Session: Metrics for Model Evaluation

  • Common metrics (FID, IS)
  • Qualitative vs. quantitative assessment

  Afternoon Session: Practical Evaluation Techniques

  • Peer review of generated outputs
  • Hands-on: Using evaluation metrics on course projects

• Day 8: Future Trends and Capstone Project

  Morning Session: The Future of Generative AI

  • Emerging trends and research directions
  • Discussion on responsible AI development

  Afternoon Session: Capstone Project Presentations

  • Participants present their projects
  • Feedback and discussion
  • Course wrap-up and Q&A