Generative Models for Text and Images - Development and Evaluation

Watch this 58-minute conference talk where Dr. Akshay Chaudhari from Stanford University explores the development and evaluation of generative models for both text and images in healthcare applications. Learn about cutting-edge approaches to creating AI systems that can generate meaningful textual and visual content, with particular emphasis on medical imaging and healthcare data contexts. Discover how self-supervised and representation learning techniques are being applied to multi-modal deep learning systems that integrate vision, language, and medical records data. Gain insights into the methodologies used to assess and validate generative models, understanding both their capabilities and limitations in healthcare settings. Explore the intersection of machine intelligence and medical imaging, including practical considerations for implementing generative AI in clinical environments. The presentation covers current research developments in the field, drawing from work conducted at Stanford's Machine Intelligence in Medical Imaging research group and the Stanford AIMI Center, providing valuable perspectives on the future of AI-generated content in healthcare applications.