Register here
Embodied artificial intelligence (EAI) marks a shift in how intelligent systems are designed and experienced. Rather than operating purely in digital environments, EAI integrates AI into physical or virtual bodies, enabling systems to perceive, learn, and act within real-world contexts. This opens new possibilities for multimedia technologies that are not only interactive, but deeply responsive to human behaviour and environments. This workshop will explore how EAI can support human-centred multimedia applications by combining multimodal perception, adaptive learning, and collaborative capabilities. Participants will engage with emerging concepts such as world models, closed-loop control, and real and virtual training datasets, while examining how embodied systems can operate in dynamic, unstructured environments. The session will also address practical implications across sectors including healthcare, smart manufacturing, logistics, and home automation. Bringing together researchers, industry experts, and policymakers, the workshop offers a space to reflect on the technical foundations, standardization needs, and future directions of embodied AI in multimedia systems.
This special session aims to formally introduce the newly established ITU Focus Group on Embodied AI for Multimedia Technologies and present its overall vision, scope, and planned activities.
This session explores the latest technological breakthroughs and global development trends in embodied AI, including hardware platforms, multimodal models, and real-world applications. The session aims to provide a comprehensive overview of the state of the art and highlight how embodied AI is expanding into industrial, service, and consumer scenarios.
This panel aims to examine the transition from R&D prototypes to scalable real-world applications, with a specific focus on recent deployments in industrial manufacturing, logistics, and transportation. The discussion will highlight practical use cases of EAI entering the workforce, addressing both technological advancements and the challenges of large-scale commercialization.
As embodied AI technologies rapidly evolve, there is a growing need for reliable evaluation frameworks, standards, and responsible governance mechanisms. This session will examine:
- Benchmarking and evaluation methodologies for embodied AI systems
- Standardization needs for multimodal models, interfaces, and system integration
- Data governance, EAI safety, and ethical considerations.
This session will focus on the transition from simulation-based validation to real-world deployment, exploring how to establish a comprehensive, fair, comparable, and reproducible benchmark testing framework for physical AI technology and relevant systems.
