Brief

Researchers have developed a new imitation learning method for robots that utilizes scene graphs to enhance their understanding of spatial and temporal context. This approach helps robots retain relevant historical information and reason over extended task horizons, particularly in environments with partial observability. Experiments in simulated and real-world manipulation tasks showed significant improvements in policy performance and generalization capabilities. AI

A scoping review of 52 studies on scene graphs in surgery reveals a significant increase in research, with a notable shift towards foundation and generative AI models. However, a 'data divide' persists, with most research using real-world endoscopic video while external operating room modeling relies on simulations. The review proposes a new 'Validation Trinity' framework to address the gap between current computer vision metrics and the needs for clinical validation of these neuro-symbolic AI systems. AI