An Introduction to Diffusion Models in Generative AI

New

With the increase in AI-generated imagery using models such as Dall-E, Midjourney and Sora and research applications such as AlphaFold, there has been a surge in workflows incorporating models like Stable Diffusion. These models have potential in research applications including drug discovery, weather forecasting, synthetic speech and medical imaging.

The aim of the session will be to equip you with knowledge of how generative AI and diffusion models work and to share an overview of research applications. The workshop will include short talks from researchers already deploying diffusion models in their research.

Much of the workshop content is conceptual and high-level, and by the end of the day participants will have a firm grasp on how diffusion models work. We won’t be coding during the session, but will share code with you for you to work with after the session.

• Postgraduate students and research staff

• To get the most out of the code, and the technical content of the workshop, we recommend that you have some exposure to the following (but please note, that this is not a requirement to attend):

Probability Deep learning Computer vision Language models PyTorch

Number of sessions: 2

• Presentations, demonstrations, group discussion and practicals

• Python will need to be downloaded prior to the course

Please note that this is a full day course and participants should book both sessions. Refreshments and lunch will be provided, please add any dietary requirements to the special requirements section.

• Full day course

• Once or twice a term

• AI and Large Language Models Workshop
• An Introduction to Docker Workshop
• Packaging and Publishing Python Code for Research workshop
• Hands On AI Workshop
• LLM Hands on Workshop