ISD8 Tutorial Report: Cognitively Inspired Reasoning for Reactive Robotics - From Image Schemas to Knowledge Enrichment

This report details a tutorial on image-schematic analysis and functional object detection presented at the conference. The tutorial bridges theoretical foundations with practical applications, introducing participants to a novel framework for analyzing video content through the lens of image schemas. Through hands-on demonstrations, participants engaged with an interactive system that processes video sequences to generate rich semantic representations. The system produces multi-layered output including knowledge graphs of image schematic activations, natural language descriptions, temporal event annotations, and automatically curated storyboards with generated captions. The theoretical underpinnings of image-schematic segmentation and functional parts detection were covered, but the tutorial focused on enabling participants to effectively utilize these tools for their own research and applications. The tutorial successfully demonstrated how image-schematic analysis can be practically applied to video understanding tasks, while gathering insights from participants about desired future functionalities. This approach of combining theoretical instruction with hands-on experimentation proved effective for both teaching the technology and collecting user requirements for future development.