UET-Headpose
收藏资源简介:
UET-Headpose数据集是由越南河内国家大学下属工程技术大学人工智能实验室创建的,旨在通过使用绝对方向传感器而非深度相机来收集头部姿态图像,以降低成本和简化设置。该数据集包含12,848张图像,覆盖了从-179到179度的全方位yaw角。数据集的创建过程涉及使用传感器和监控摄像头,以及手动标注图像。UET-Headpose数据集主要应用于人机交互和顾客行为分析等领域,旨在解决现有数据集在处理大角度数据时的不足。
The UET-Headpose dataset was created by the Artificial Intelligence Laboratory at the University of Engineering and Technology, Vietnam National University, Hanoi. It aims to collect head pose images using absolute orientation sensors rather than depth cameras, thereby reducing costs and simplifying the experimental setup. This dataset contains 12,848 images, covering full-range yaw angles ranging from -179° to 179°. The construction of this dataset involved the utilization of sensors and surveillance cameras, alongside manual image annotation. The UET-Headpose dataset is primarily applied in fields such as human-computer interaction and customer behavior analysis, aiming to address the limitations of existing datasets when handling large-angle pose data.




