{"url":"/dataset/berst","name":"BERSt","full_name":"Basic Emotion Random phrase Shouts","description_markdown":"BERSt Dataset\r\n\r\nWe release the BERSt Dataset for various speech recognition tasks including Automatic Speech Recognition (ASR) and Speech Emotion Recogniton (SER)\r\nOverview\r\n\r\n    4526 single phrase recordings (~3.75h)\r\n    98 professional actors\r\n    19 phone positions\r\n    7 emotion classes\r\n    3 vocal intensity levels\r\n    varied regional and non-native English accents\r\n    nonsense phrases covering all English Phonemes\r\n\r\nData collection\r\n\r\nThe BERSt dataset represents data collected in home envrionments using various smartphone microphones (phone model available as metadata) Participants were around the globe and represent varying regional accents in English: UK, Canada, USA (multi-state), Australia, including a subset of the data that is non-native English speakers including: French, Russian, Hindi etc. The data includes 13 non-sense phrases for use cases robust to lingustic context and high surprisal. Partipants were prompted to speak, raise their voice and shout each phrase while moving their phone to various distances and locations in their home, as well as with various obstructions to the microphone, e.g. in a backpack\r\n\r\nBaseline results of various state-of-the-art methods for ASR and SER show that this dataset remains a challenging task, and we encourage researchers to use this data to fine-tune and benchmark their models in these difficult condition representing possible real world situations\r\n\r\nAffect annotations are those provided to the actors, they have not been validated through perception The speech annotations, however, has been checked and adjusted to mistakes in the speech.\r\nData splits and organisation\r\n\r\nFor each phone position and phrase, the actors provided a single recording for the three vocal intensity levels, these raw audio files are available\r\n\r\nMeta-data in csv format corresponds to the files split per utterance with noise and silence before and after speech removed, found inside clean_clips for each data splits\r\n\r\nWe provide a test, train and validation split\r\n\r\nThere is no speaker cross-over between splits, the train and validation sets each contain 10 speakers not seen in the training set\r\n\r\nMetadata Details\r\n\r\n    actor count\r\n        98\r\n    Gender counts\r\n        Woman: 61\r\n        Man: 34\r\n        Non-Binary: 1\r\n        Prefer not to disclose 2\r\n    Current daily language counts\r\n        English: 95\r\n        Norwegian: 1\r\n        Russian: 1\r\n        French: 1\r\n    First language counts\r\n        English: 75\r\n        Non English: 23\r\n            Spanish: 6\r\n            French: 3\r\n            Portuguese: 3\r\n            Chinese: 2\r\n            Norwegian: 1\r\n            Mandarin: 1\r\n            Tagalog: 1\r\n            Italian: 1\r\n            Hungarian: 1\r\n            Russian: 1\r\n            Hindi: 1\r\n            Swahili: 1\r\n            Croatian: 1\r\nPre-split Data counts\r\n    Emotion counts\r\n        fear: 236\r\n        neutral: 234\r\n        disgust: 232\r\n        joy: 224\r\n        anger: 223\r\n        surprise: 210\r\n        sadness: 201\r\n    Distance counts:\r\n        Near body: 627\r\n        1-2m away: 324\r\n        Other side of room: 316\r\n        Outside of room: 293","description_withheld":null,"homepage":"https://huggingface.co/datasets/Rosie-Lab/BERSt","introduced_date":"2025-04-30","introduced_date_note":null,"introduced_by":{"paper":"/paper/bersting-at-the-screams-a-benchmark-for","title":"BERSting at the Screams: A Benchmark for Distanced, Emotional and Shouted Speech Recognition","first_author":"Paige Tuttösí","url":null},"license":{"name":"Creative Commons Attribution 4.0","url":null},"modalities":[{"name":"Audio","url":"/datasets/modality/audio"}],"tasks":[{"name":"Speech Recognition","url":"/task/speech-recognition","datasets_with_task":"/datasets/task/speech-recognition"},{"name":"Emotion Recognition","url":"/task/emotion-recognition","datasets_with_task":"/datasets/task/emotion-recognition"},{"name":"Automatic Speech Recognition","url":"/task/automatic-speech-recognition-2","datasets_with_task":"/datasets/task/automatic-speech-recognition-2"},{"name":"Speech Emotion Recognition","url":"/task/speech-emotion-recognition","datasets_with_task":"/datasets/task/speech-emotion-recognition"},{"name":"Distant Speech Recognition","url":"/task/distant-speech-recognition","datasets_with_task":"/datasets/task/distant-speech-recognition"},{"name":"Perceptual Distance","url":"/task/perceptual-distance","datasets_with_task":"/datasets/task/perceptual-distance"}],"languages":[{"name":"English","url":"/datasets/language/english"}],"variants":["BERSt"],"data_loaders":[],"num_papers_in_archive":1,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[{"leaderboard":"/sota/speech-emotion-recognition-on-berst","task":"Speech Emotion Recognition","dataset_variant":"BERSt","rows":3,"metrics":["Unweighted Accuracy (UA)","Weighted Accuracy (WA)"],"first_row_in_archive_order":{"model":"DAWN-hidden-SVM","paper":"/paper/bersting-at-the-screams-a-benchmark-for","metrics":{"Unweighted Accuracy (UA)":"32.1","Weighted Accuracy (WA)":"32.2"},"code_links":[{"title":"myHaiven/data-collection","url":"https://github.com/myHaiven/data-collection"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"}],"papers_with_a_benchmark_row":[{"paper":"/paper/bersting-at-the-screams-a-benchmark-for","title":"BERSting at the Screams: A Benchmark for Distanced, Emotional and Shouted Speech Recognition","date":"2025-04-30","rows_on_this_dataset":3,"code_links":1,"syntology":null}],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":0,"samples_harvested":0,"samples_ran":0,"samples_unverified":0,"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."}