• ES Español
• EN English

5.56. Advanced Generative AI Models (Mandatory)

5.56.1. Justification ↑ Back to top

This course explores modern generative models (GANs, Diffusion Models, LLMs) for multimodal content creation, covering both mathematical foundations and practical applications with ethical considerations for synthetic media.

5.56.2. Generales Goals ↑ Back to top

1. Implement image/text/video generation pipelines
2. Design AI-assisted editing systems
3. Evaluate synthetic content risks

5.56.3. Contribution to Outcomes ↑ Back to top

AG-C09) Design and Development of Solutions: Designs, implements, and evaluates solutions for complex computing problems. (Assessment)

AG-C11) Use of Tools: Applies modern computing tools in problem solving. (Assessment)

5.56.4. Content ↑ Back to top

5.56.4.1. Generative Fundamentals (12 hours) [Skills AG-C09,AG-C11] ↑ Back to top

Bibliography: (Goodfellow et al., 2014)

Topics

1. GANs theory (minimax, Wasserstein)
2. Autoregressive models (PixelCNN, Transformers)
3. Evaluation metrics (FID, Inception Score)

Learning Outcomes

1. Train basic GANs [Usage]
2. Analyze generation failure modes [Assessment]

5.56.4.2. Diffusion Models (24 hours) [Skills AG-C09,AG-C11] ↑ Back to top

Bibliography: (Sohl-Dickstein et al., 2023; Rombach et al., 2022)

Topics

1. Diffusion processes (forward/reverse)
2. Stable Diffusion and latent architectures
3. Control techniques (ControlNet, LoRA)

Learning Outcomes

1. Implement diffusion samplers [Usage]
2. Fine-tune for domain-specific generation [Assessment]

5.56.4.3. Synthetic Media Ethics (8 hours) [Skills AG-C09,AG-C11] ↑ Back to top

Bibliography: (Sohl-Dickstein et al., 2023)

Topics

1. Deepfakes and misinformation
2. Copyright in generated content
3. Authenticity marking (C2PA, watermarking)

Learning Outcomes

1. Detect synthetic content [Familiarity]
2. Implement traceability systems [Usage]

5.56.5. Bibliography ↑ Back to top

Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A., and Bengio, Y. (2014). Generative adversarial networks. In NeurIPS.

Sohl-Dickstein, J., Song, Y., and Ermon, S. (2023). Diffusion Models: A Comprehensive Practical Guide. AI Press.

Rombach, R., Blattmann, A., Lorenz, D., Esser, P., and Ommer, B. (2022). High-resolution image synthesis with latent diffusion models. CVPR.