AffectAI-Capture Protocol for Small-Group Meeting Research
AffectAI-Capture has been created by researchers as a reproducible method for gathering synchronized multimodal data during interactions resembling four-person meetings. This protocol combines eye tracking, wearable physiological sensors, close-talk and ambient audio, multi-angle video, event logging, and structured self-reports. Sessions are organized into fixed task blocks grounded in established group-interaction paradigms. A unified event timeline guides data collection and post-processing, ensuring standardized outputs. The rationale behind the protocol's design, synchronization approach, data structuring, and practical compromises are detailed. While pilot tests have confirmed audio quality and video synchronization, full protocol sessions with participants are still in progress. This framework offers a reproducible architecture for research in affective, behavioral, and meeting analytics.
Key facts
- AffectAI-Capture is a protocol for four-person meeting-like interactions.
- It combines eye tracking, wearable physiology, audio, video, event logging, and self-report.
- Sessions use fixed task blocks from established group-interaction paradigms.
- Data is organized around a single authoritative event timeline.
- Pilot validation of audio and video synchronization is complete.
- Full protocol sessions with participants are ongoing.
- The protocol is designed for affective, behavioral, and meeting-analytics research.
- The work is published on arXiv under Computer Science > Human-Computer Interaction.
Entities
Institutions
- arXiv