Quazitron420/video-dataset-pre-test-colab

这个数据集包含了从注释视频片段中提取的视频帧，以及相应的注释、转录和视频剪辑。数据集的结构包括按角色分组的提取帧（开始、中间、结束），每个注释区间的视频片段，原始的JSON注释文件，每个片段的转录文件，以及一个映射帧、注释、视频片段和片段转录的dataset.csv文件。数据集共有42帧，14个片段，7个独特的标签。

This dataset contains video frames extracted from annotated video segments, along with annotations, transcriptions, and corresponding video clips. The dataset structure includes extracted frames grouped by role (start, middle, end), video clips for each annotation interval, original JSON annotation files, transcription files for each segment, and a dataset.csv file mapping frames, annotations, video clips, and segment transcriptions. There are a total of 42 frames, 14 segments, and 7 unique labels.