New COSI-Lab dataset models multi-perspective social intentions

Researchers have introduced COSI-Lab, a novel dataset and framework designed to model multi-perspective social intentions in complex environments. This dataset captures interactions from a scientific workshop, focusing on Apparent Intent Inference (AII) by considering observers' interpretative tendencies. COSI-Lab aims to equip future intelligent systems with better subjective perception capabilities by treating the multiplicity of intentions as explainable, perspective-driven reasoning rather than mere label noise. The project includes a new annotation process for AII, quantitative and qualitative analyses of intent narratives, benchmark tasks, and multimodal data for behavior analysis. AI

IMPACT This research could lead to AI systems that better understand and interpret complex human social dynamics and intentions.

RANK_REASON The cluster describes a new academic paper and dataset release. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• alphaXiv
• Apparent Intent Inference
• arXiv
• CatalyzeX
• CORE Recommender
• COSI-Lab
• DagsHub
• Gotit.pub
• Hugging Face
• Influence Flower
• ScienceCast

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →