icomgpu/rccai

这个数据集包含从注释视频片段中提取的视频帧，以及注释、转录和相应的视频剪辑。数据集的结构包括提取的帧（每个片段的第一帧）、每个注释间隔的视频剪辑、原始JSON注释、转录文件（`full_transcription.txt` + 每个片段）以及一个映射帧、视频剪辑、转录和元数据（主题、类别、子类别、动作描述、上下文和评论）的CSV文件。数据集的特征包括图像（JPEG格式的提取视频帧）、视频片段路径、转录文本、主题（例如“Учитель”、“Группа учеников”）、类别（例如“Педагогические действия”、“Коммуникативные модальности”）、子类别（可能包含多个以“;”分隔的值）、动作描述以及上下文和附加评论。数据集的基本统计信息包括54帧、18个片段和1个唯一标签。

This dataset contains video frames extracted from annotated video segments, along with annotations, transcriptions, and corresponding video clips. The dataset structure includes extracted frames (first frame from each segment), video clips for each annotation interval, original JSON annotation, transcription files (`full_transcription.txt` + per segment), and a CSV file mapping between frames, video clips, transcriptions, and metadata (subject, category, subcategory, action_description, context_and_comments). The dataset features include image (extracted video frame in JPEG format), video_segment (path to the corresponding video clip file), transcription (text transcription of the audio segment), subject (e.g., "Учитель", "Группа учеников"), category (e.g., "Педагогические действия", "Коммуникативные модальности"), subcategory (may contain multiple values separated by ";"), action_description (description of the action), and context_and_comments (context and additional comments). The dataset statistics include 54 frames, 18 segments, and 1 unique label.