ASD1 dataset

该数据集包含了10位健康的法语母语者讲述的77个句子，这些语音数据经过了同步录制、去噪和与文本及视频对齐处理。此外，图像以每秒50帧的速度捕获，具有136×136像素的分辨率和20毫秒的时间分辨率。尽管部分发言者可能无法看到明显的喉部，但他们的发音动作仍具有宝贵价值。该数据集的规模为77个句子，旨在支持从RTMRI数据中进行语音合成和文本预测的任务。

This dataset contains 77 sentences spoken by 10 healthy French native speakers. The accompanying audio data was synchronously recorded, denoised, and aligned with corresponding text and video materials. Additionally, the imaging data was captured at 50 frames per second, with a resolution of 136×136 pixels and a temporal resolution of 20 milliseconds. Although some speakers may not have clearly visible larynges, their articulatory movements still hold significant research value. With a total of 77 sentences, this dataset is designed to support speech synthesis and text prediction tasks based on RTMRI data.