Nursing Operation Gesture Video Dataset

In the current healthcare field, the standardization and precise evaluation of nursing operations are crucial. However, the actions of nurses during operations are complex and varied, and traditional manual-based training methods no longer meet the needs for efficient and personalized teaching. At the same time, existing solutions such as purely text-based guidance or limited demonstration videos fail to fully capture the detailed movements and consistent evaluation standards. This dataset aims to provide high-resolution videos of nursing operations to support the automatic recognition and evaluation of complex actions, thereby improving training efficiency and the quality of nursing services.Data collection uses high-definition camera equipment, and standardized operation demonstrations are performed by professional nurses in a simulated ward environment. The data is annotated through multiple rounds to ensure clarity and accuracy of action steps, with the annotation team comprising individuals with professional medical backgrounds, totaling 30 members. Data preprocessing includes action segmentation, interframe denoising, and video compression to ensure consistency and reduce storage space. The data is stored in MP4 format and organized for easy indexing and retrieval.Regarding data quality, the dataset ensures annotation precision, with action recognition accuracy exceeding 95%. Technological innovations include the introduction of new annotation methods and automated quality assessment tools, enhancing annotation consistency. The application value lies in the accurate recognition of nursing actions, improved nurse training effectiveness, with wider and more detailed action coverage compared to similar datasets. The dataset's uniqueness is in its high-quality annotation and applicability, with good extensibility, suitable for different global nursing standards, and offering high versatility.