LAION-BVD: A 10-Million-Hour Open Video Dataset for Multimodal Pre-training

Authors:Andreas Hochlehnert, Marianna Nezhurina, Mehdi Cherti, Andrej Radonjic, Thaddäus Wiedemer, Christoph Schuhmann, Romain Beaumont, Wieland Brendel, Bernhard Schölkopf, A. Sophia Koepke, Jenia Jitsev, Matthias Bethge

Abstract:We present LAION-BVD, a large-scale open video dataset for multimodal learning, which contains 1.3B platform-specific video URLs collected from CommonCrawl. From these, we download 80M videos with a total duration of 10 million hours. The dataset is designed for multimodal pre-training across the video, audio, and image modalities. Using content-aware scene detection, we extract clips for which we synthetically generate video and audio captions. Models trained on these data achieve competitive performance on standard video-text and audio-text benchmarks, with consistent improvements as training or model scale increases. Additionally, we explore video frames as an alternative source of image-text data by extracting scene-changing frames. These frames exhibit a visual distribution distinct from standard web image corpora, and models trained on this dataset achieve strong image-text retrieval performance. We release LAION-BVD to the research community. It significantly expands open access to multimodal videos at an unprecedented scale.

Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)

arXiv:2608.24845 [cs.CV]

(or arXiv:2608.24845v1 [cs.CV] for this version)

https://doi.org/10.48550/arXiv.2608.24845 arXiv-issued DOI via DataCite (pending registration)