QualiSpeech

该数据集名为QualiSpeech，是一个全面的低级别语音质量评估数据集，涵盖了11个关键方面，并附有详细的自然语言评论，这比传统的数值评分能提供更丰富的洞察。此外，该数据集在七个维度上进行了标注，并为四个方面提供了简洁的描述，开创性地在语音质量评估中使用了自然语言描述。该数据集规模宏大，包含了多样化的场景和方面，并配有详尽的注释，其任务是进行语音质量评估。

This dataset is named QualiSpeech, a comprehensive low-level speech quality assessment dataset covering 11 key aspects and accompanied by detailed natural language reviews, which provides richer insights than traditional numerical scores. Additionally, the dataset is annotated across seven dimensions, with concise descriptions provided for four aspects, pioneering the application of natural language descriptions in speech quality assessment. Boasting a large scale, it encompasses diverse scenarios and aspects with exhaustive annotations, and its targeted task is speech quality assessment.