Co-working Space Personnel Collaboration Image Dataset

In modern office environments, maximizing employee collaboration efficiency has become a key challenge for enterprises. Current office behavior analysis relies heavily on limited sensor data, lacking comprehensive observation from a visual perspective, leading to inaccuracies and incomplete identification of collaboration behaviors. The Co-working Space Personnel Collaboration Image Dataset aims to address the technical challenge of collaboration detail identification through rich image data, providing effective optimization suggestions. During data collection, high-definition cameras are set up in different positions to capture multi-angle images of people collaborating, ensuring data comprehensiveness and coverage. Multi-round annotation and consistency checks are carried out by a 30-person team with experience in behavioral science and data annotation, further ensuring annotation accuracy and consistency. Data preprocessing includes steps such as image cropping and color correction, and is organized and stored in JPG format for easy parsing and processing.