{"url":"/dataset/youtube-8m","name":"YouTube-8M","full_name":null,"description_markdown":"The **YouTube-8M** dataset is a large scale video dataset, which includes more than 7 million videos with 4716 classes labeled by the annotation system. The dataset consists of three parts: training set, validate set, and test set. In the training set, each class contains at least 100 training videos. Features of these videos are extracted by the state-of-the-art popular pre-trained models and released for public use. Each video contains audio and visual modality. Based on the visual information, videos are divided into 24 topics, such as sports, game, arts & entertainment, etc\r\n\r\nSource: [Audio-Visual Embedding for Cross-Modal Music Video Retrieval through Supervised Deep CCA](https://arxiv.org/abs/1908.03744)","description_withheld":null,"homepage":"https://research.google.com/youtube8m/","introduced_date":"2016-01-01","introduced_date_note":null,"introduced_by":{"paper":"/paper/youtube-8m-a-large-scale-video-classification","title":"YouTube-8M: A Large-Scale Video Classification Benchmark","first_author":"Sami Abu-El-Haija","url":null},"license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"modalities":[{"name":"Videos","url":"/datasets/modality/videos"}],"tasks":[{"name":"Video Prediction","url":"/task/video-prediction","datasets_with_task":"/datasets/task/video-prediction"},{"name":"Video Classification","url":"/task/video-classification","datasets_with_task":"/datasets/task/video-classification"}],"languages":[],"variants":["YouTube-8M"],"data_loaders":[],"num_papers_in_archive":147,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[{"leaderboard":"/sota/video-classification-on-youtube-8m","task":"Video Classification","dataset_variant":"YouTube-8M","rows":3,"metrics":["Hit@1","PERR","Hit@5","Global Average Precision","mAP"],"first_row_in_archive_order":{"model":"DCGN (self-attention graph pooling)","paper":"/paper/190600377","metrics":{"Hit@1":"87.7"},"code_links":[]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/video-prediction-on-youtube-8m","task":"Video Prediction","dataset_variant":"YouTube-8M","rows":1,"metrics":["Average PSNR"],"first_row_in_archive_order":{"model":"SDCNet","paper":"/paper/sdc-net-video-prediction-using-spatially","metrics":{"Average PSNR":"37.15"},"code_links":[{"title":"NVIDIA/semantic-segmentation","url":"https://github.com/NVIDIA/semantic-segmentation/tree/sdcnet/sdcnet"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"}],"papers_with_a_benchmark_row":[{"paper":"/paper/190600377","title":"Hierarchical Video Frame Sequence Representation with Deep Convolutional Graph Network","date":"2019-06-02","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/efficient-video-classification-using-fewer","title":"Efficient Video Classification Using Fewer Frames","date":"2019-02-27","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/sdc-net-video-prediction-using-spatially","title":"SDC-Net: Video prediction using spatially-displaced convolution","date":"2018-09-01","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/youtube-8m-a-large-scale-video-classification","title":"YouTube-8M: A Large-Scale Video Classification Benchmark","date":"2016-09-27","rows_on_this_dataset":1,"code_links":7,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":9,"samples_ran":1,"samples_unverified":8,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}}],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":1,"samples_harvested":9,"samples_ran":1,"samples_unverified":8,"pointer_only_for_licence":0,"papers_with_no_sample_that_ran":0,"note":"the per-paper counts above, summed; not a rate"},"papers_note":"The archive never published its papers-using-dataset list; these are papers with a leaderboard row on this dataset's benchmarks."}