AWCV-100K-UCAS2024
收藏资源简介:
AWCV-100K-UCAS2024是由中国科学院大学团队创建的视频空中手写汉字数据集,旨在解决现有空中手写技术依赖复杂传感器和缺乏视频数据的问题。该数据集包含102,688个视频,共计880万帧,涵盖了GB2312-80一级字库中的3,755个汉字,覆盖了日常使用汉字的99.7%。数据集通过普通RGB摄像头采集,适用于多种现实场景,如智能设备、虚拟现实等。数据集的创建过程包括使用空中手写平台采集数据,参与者被要求自然书写汉字。该数据集的应用领域包括人机交互、智能对话系统等,旨在提升空中手写识别的准确性和实用性。
AWCV-100K-UCAS2024 is a video-based air-writing Chinese character dataset developed by the team from the University of Chinese Academy of Sciences (UCAS) in 2024. It is designed to solve the problems that existing air-writing technologies rely on complex sensors and face the shortage of available video data. This dataset includes 102,688 video clips with a total of 8.8 million frames, covering 3,755 Chinese characters from the GB2312-80 first-level character set, which accounts for 99.7% of daily-used Chinese characters. Collected via ordinary RGB cameras, this dataset is suitable for multiple real-world scenarios such as smart devices and virtual reality (VR). The dataset construction process uses an air-writing platform for data collection, where participants were asked to write Chinese characters naturally. Its application fields cover human-computer interaction, intelligent dialogue systems and more, aiming to improve the accuracy and practicality of air-writing recognition.




