MLLMGUARD: A Multi-dimensional Safety Evaluation Suite for Multimodal Large Language Models

This 26-minute presentation from the University of Central Florida introduces MLLMGUARD, a comprehensive multi-dimensional safety evaluation suite designed specifically for Multimodal Large Language Models. Explore how this framework assesses potential risks and vulnerabilities in systems that process both text and visual inputs. Learn about the methodology, benchmarks, and findings that help identify safety concerns in these increasingly prevalent AI systems. Access the complete slide deck to understand the technical details and implications for developing more secure multimodal AI technologies.