msnowchanj/CORA
收藏官方服务：
资源简介：
这是一个多模态音频-文本数据集，包含音频文件及相关文本信息，如原始字幕、关键词、陈述、问题、命令和间接描述。数据集用于测试，共有3367个样本，涉及多个特征字段，支持音频识别、生成或评估任务，可能应用于自然语言处理与音频处理的交叉领域。
This is a multimodal audio-text dataset containing audio files and associated textual information such as original captions, key phrases, statements, questions, commands, and indirect descriptions. The dataset is designed for testing, with 3367 samples, and includes multiple feature fields to support audio recognition, generation, or evaluation tasks, potentially applied in cross-domain areas of natural language processing and audio processing.
提供机构：
msnowchanj
