BVD-V-55M-CC — the Creative Commons subset

Every video in laion/BVD-V-55M-URLs whose YouTube licence is creativeCommon, in the same format as the original.

The licence is not part of BVD, so it was resolved for the whole corpus through the YouTube Data API: all 2,119,224 source videos, one videos.list(part=status) call per fifty ids. 15,658 came back creativeCommon, 0.74%. This is a complete census rather than a sample, so the subset is exhaustive: no CC video in BVD-V-55M is missing from it.

Licences change. A creator can switch a video back to the standard YouTube licence at any time, so treat this as a snapshot taken 2026-09-25 and re-check before relying on it.

Repository structure

Identical to the parent dataset, so anything written against BVD-V-55M-URLs works here unchanged.

videos/          # one row per unique source video, all 27 original columns
  0000.parquet
  ...
clips/           # one row per scene clip, all 10 original columns
  0000.parquet
  ...

Join the two on id, exactly as in the parent dataset.

Loading

import datasets

videos = datasets.load_dataset("datachain/BVD-V-55M-CC", "videos")
clips  = datasets.load_dataset("datachain/BVD-V-55M-CC", "clips")

Notes

No media is included, same as the parent dataset: only URLs, timestamps and text annotations. Clips can be reproduced from webpage_url plus start_time and end_time with yt-dlp and ffmpeg.

A creativeCommon licence is what YouTube reports for the video. It says the uploader marked it reusable; it does not verify that they held the rights to do so.

Source

• Parent dataset: https://huggingface.co/datasets/laion/BVD-V-55M-URLs
• Project page: https://projects.laion.ai/bvd/
• Paper: https://arxiv.org/abs/2608.24845