◕ Track 1: Foundation Models for Embodied Intelligence

· Architecture Design for Robotic Foundation Models 

· World Models for Embodied Agents

· Scaling Laws and Emergent Abilities in Embodied Systems

· Efficient and Low-Latency Inference for Real-World Deployment

· Theoretical Foundations of Embodied Intelligence

◕ Track 2: Learning, Adaptation, and Generalization in Embodied Systems

·Reinforcement Learning and Imitation Learning from Multimodal Data

· Zero/Few-Shot and Long-tail Learning for Embodied Agents

· Continual/Lifelong Learning in Dynamic Environments

· Embodied Chain-of-Thought (CoT) Reasoning with LLMs

· Learning from Physical Interaction, Correction, and Human Feedback

◕ Track 3: Multimodal Perception and Interaction

· Multimodal Perception and Fusion 

· 3D Scene Understanding and Semantic Mapping

· Affordance Prediction and Functional Object Recognition

· Egocentric Vision and Embodied Vision-Language Models

· Multimodal Sensor Data Fusion

· Localization, Navigation and Controlling

◕ Track 4: Systems and Real-World Applications

· Multi-Agent Collaboration Systems 

· Human-Robot Interaction

· Robotic Systems Integration and Deployment 

· Edge Computing, End-Cloud Collaboration and Communication Technologies

· Embodied AI in Manufacturing, Logistics, Agriculture, and etc

◕ Track 5: Safety, Interpretability and Ethics

· Alignment and Safety in Physical Systems

· Bias, Fairness, and Transparency in Embodied AI

· Interpretability and Explainability of Embodied Models

· Data Privacy and Security in Human-Centric Environments

· Ethical Human-robot Interaction and Societal Impact

◕ Track 6: Simulation, Datasets and Environments

· Large-scale Embodied Data Collection and Curation

· Image and Video Synthesis and Generation

· Benchmarks and Metrics for Evaluating Embodied Intelligence

· Human-in-the-Loop Data Collection and Annotation Frameworks 

· Standardization of Embodied AI Evaluation Protocols