ENIGMA-360
收藏资源简介:
ENIGMA-360是由卡塔尼亚大学IPLAB实验室与Next Vision公司联合开发的工业场景多视角行为理解数据集,包含360条同步采集的自我中心与外部中心视频对。该数据集以真实工业实验室为背景,记录了工人执行维护任务的完整流程,视频总时长未明确但覆盖两种视角各180段,并标注了时间动作分段、关键步骤识别等时空标签。数据通过可穿戴设备与固定摄像头同步采集,辅以工业专家设计的标准化流程。其核心目标是推动工业环境下人机交互检测、安全合规监测等AI辅助系统的研发,填补了真实工业场景多视角行为数据的空白。
ENIGMA-360 is a multi-view behavior understanding dataset for industrial scenarios, jointly developed by the IPLAB Laboratory of the University of Catania and Next Vision Company. It contains 360 synchronized pairs of egocentric and exocentric videos. Based on a real industrial laboratory setting, the dataset records the complete workflow of workers performing maintenance tasks. The total runtime of the videos is not specified, but the dataset covers 180 segments for each of the two perspectives. It is annotated with spatiotemporal labels including temporal action segmentation and key step recognition. The data is synchronously collected via wearable devices and fixed cameras, supplemented by standardized workflows designed by industrial experts. Its core objective is to promote the research and development of AI-assisted systems such as human-computer interaction detection and safety compliance monitoring in industrial environments, filling the gap in multi-view behavior data for real industrial scenarios.
- 1ENIGMA-360: An Ego-Exo Dataset for Human Behavior Understanding in Industrial Scenarios卡塔尼亚大学·数学与计算机科学系IPLAB实验室; Next Vision s.r.l. - 卡塔尼亚大学衍生公司 · 2026年



