巴西手语数据集

巴西手语数据集是由圣保罗州立大学创建，旨在通过深度学习和图像处理技术提高手语识别的效率。该数据集包含14个日常使用的手语词汇，由不同智能手机在RGB颜色系统下录制，总计84个视频。数据集的创建过程中，考虑了帧率、背景类型、预处理和数据增强等因素，使用卷积神经网络和对象检测技术进行图像分类。该数据集主要应用于解决听障人士的沟通问题，通过提供高质量的手语识别模型，增强听障社区的交流能力。

The Brazilian Sign Language Dataset was created by São Paulo State University, aiming to improve the efficiency of sign language recognition via deep learning and image processing technologies. This dataset contains 14 daily-used sign language vocabulary items, recorded by different smartphones under the RGB color system, with a total of 84 videos. During the dataset construction, factors such as frame rate, background types, preprocessing and data augmentation were taken into consideration. Convolutional neural networks (CNNs) and object detection technologies were used for image classification. This dataset is mainly applied to solve the communication problems of the hearing-impaired, and enhance the communication ability of the hearing-impaired community by providing high-quality sign language recognition models.