Moment-Video: Annotated Video Dataset for Question Answering

A video dataset with annotations for category, subclass, and question-answer pairs. The dataset was created by VisionXLab and was last updated on 2026-05-26. It contains video files and corresponding annotation files in JSON and CSV formats.

Use Cases

• Train video question answering models based on the provided question-answer annotations.
• Classify videos into categories and subclasses based on the annotated labels.
• Benchmark multimodal reasoning systems on tasks involving video content.

Strengths

• Annotations include category, subclass, and question-answer pairs, which suggests structured labels for analysis.
• Data is provided in both JSON and CSV formats, offering flexibility for different processing workflows.

Limitations

• The total number of videos, rows, and specific column definitions are unknown, limiting suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

VisionXLab

Freshness

Last updated 2026-05-26 14:35:46; freshness should be verified.