A 13-Million-Hour Open Video Dataset for Multimodal Pre-training

View a PDF of the paper titled LAION-BVD: A 13-Million-Hour Open Video Dataset for Multimodal Pre-training, by Andreas Hochlehnert and Marianna Nezhurina and Mehdi Cherti and Andrej Radonjic and Thadd\”aus Wiedemer and Christoph Schuhmann and Romain Beaumont and Wieland Brendel and Bernhard Sch\”olkopf and A. Sophia Koepke and Jenia Jitsev and Matthias Bethge

View PDF
HTML (experimental)

Abstract:We present LAION-BVD (LAION – Big Video Dataset), 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 13.5 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.