Dr. Prosenjit Banerjee

Dive deep into the world of AI's latest foundation models in this informative one-hour session. We'll explore the intricacies of large multimodal models (LMMs), their emergent applications, and how they revolutionize content generation. Discover the role of a mixture of experts in enhancing AI capabilities and envision the future of AI agents working alongside these powerful models.

The session covers large multimodal models (LMMs), starting with an introduction to multimodal learning and the architecture of LMMs. It highlights the importance of integrating multiple data types, explores various application scenarios, and discusses evaluation metrics and benchmarks. The session also delves into fine-grained grounding, the mixture of experts' approaches, AI content generation aligned with human intentions, and the role of AI agents in interpreting and interacting with LMMs, with real-world examples throughout.