Rapidata/sora-video-generation-aligned-words

Rapidata视频生成单词对齐数据集，包含约1500名人类评估者对AI生成视频的评价，评估依据是提示文本与视频内容的不匹配程度。该数据集基于对齐数据集，筛选出评分低于0.5的视频进行详细分析，包含类别、提示、视频、结果和用户详细信息等特征。

Rapidata Video Generation Word for Word Alignment Dataset, which includes evaluations from approximately 1500 human evaluators on AI-generated videos based on the degree of mismatch between the prompt text and the video content. The dataset is based on the alignment dataset and selects videos with scores below 0.5 for detailed analysis, containing features such as category, prompt, video, results, and user details.