Real World Engagement Attention Dataset #1c (RWEAD #1c)

RWEAD #1c is a group-behaviour video dataset acquired in an approximately 57 m² room (~10.8 m × ~7.0 m) of a university building hall, configured as an Indoor-category exhibition stand within a building hall. The sensing infrastructure comprises a single RGB camera (1920 × 1080 px, 30 fps) mounted at an elevated viewpoint, 3.5 m high and tilted 20° downward, overlooking the entire area. The positions of the camera and of the three Areas of Interest (AoIs) -- Desk, Maquette and TV -- were logged in a coordinate table. Because the hall has a glazed façade, people outside the area remain visible after exiting; an indoor event-area mask therefore separates the indoor area from the outside, and entries and exits are defined as crossings of the two door lines. The participant cohort consists of seven amateur actors (P#1–P#7) aged 20–30, all university students without formal acting training. Each participant received a script and was allotted five minutes for memorisation prior to data acquisition. Fifteen scripted recordings were acquired, five for each construct: conversion, retention, and commitment. The scripts include "positive" and "negative" cases, as well as group dynamics such as groups merging and splitting, members leaving and rejoining their group, exits and re-entries, overlapping groups, and occlusion by a structural pillar.