The dataset viewer is not available for this subset.
Exception: SplitsNotFoundError
Message: The split names could not be parsed from the dataset config.
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 286, in get_dataset_config_info
for split_generator in builder._split_generators(
~~~~~~~~~~~~~~~~~~~~~~~~~^
StreamingDownloadManager(base_path=builder.base_path, download_config=download_config)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
)
^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/webdataset/webdataset.py", line 84, in _split_generators
raise ValueError(
...<2 lines>...
)
ValueError: The TAR archives of the dataset should be in WebDataset format, but the files in the archive don't share the same prefix or the same types.
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/src/services/worker/src/worker/job_runners/config/split_names.py", line 68, in compute_split_names_from_streaming_response
for split in get_dataset_split_names(
~~~~~~~~~~~~~~~~~~~~~~~^
path=dataset,
^^^^^^^^^^^^^
config_name=config,
^^^^^^^^^^^^^^^^^^^
token=hf_token,
^^^^^^^^^^^^^^^
)
^
File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 340, in get_dataset_split_names
info = get_dataset_config_info(
path,
...<6 lines>...
**config_kwargs,
)
File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 291, in get_dataset_config_info
raise SplitsNotFoundError("The split names could not be parsed from the dataset config.") from err
datasets.inspect.SplitsNotFoundError: The split names could not be parsed from the dataset config.Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
URBAN-SED v2.0.0 (redistributed copy)
This is a redistribution of a dataset created by other people. All credit belongs to the original authors: Justin Salamon, Duncan MacConnell, Mark Cartwright, Peter Li and Juan Pablo Bello (Music and Audio Research Lab and Center for Urban Science and Progress, New York University). We did not create, modify, filter or re-annotate it.
- Original record: https://zenodo.org/records/1324404 (DOI 10.5281/zenodo.1324404)
- Original license: Creative Commons Attribution 4.0 International (CC BY 4.0), as stated on the Zenodo record. This copy keeps the same license; you must give appropriate credit to the original authors and indicate if you make changes.
- Content:
URBAN-SED_v2.0.0.tar.gz, byte-identical to the Zenodo file (6,641,129,513 bytes, md5a2d24a2148ece7c021fcc079ee87c2dc). Please check the md5 after download.
What it is
According to the README.txt shipped inside the archive, URBAN-SED is a set of 10,000 synthesised
10-second soundscapes (mono, 44.1 kHz, 16-bit WAV) with strong sound event annotations, generated with
Scaper. It totals almost 30 hours and about 50,000 annotated
events from 10 classes (air_conditioner, car_horn, children_playing, dog_bark, drilling, engine_idling,
gun_shot, jackhammer, siren, street_music). It is pre-split into 6,000 train, 2,000 validate and 2,000
test soundscapes. Annotations come as JAMS files and as simplified tab-separated text files. For
construction details read the original record and the paper below.
Source material and credits
The sound events are clips from the UrbanSound8K dataset, which in turn uses sounds from
Freesound. The archive contains FREESOUNDCREDITS.txt, which lists the
Freesound sound ids and their contributors; those contributors deserve credit too. The archive also
contains the full text of the CC BY 4.0 license (LICENSE.txt). We state the license of URBAN-SED as
given by its authors; we did not separately re-check the licenses of the individual source clips.
Why this copy exists
We mirror it so that experiments in our sound-event-detection research (project PACE) can be reproduced from one place with a fixed file. It is not an official mirror, and the original authors did not review or endorse this repository or our work. If the original record and this copy ever differ, the original record is authoritative. If you are an author and want this copy changed or removed, open a discussion on this repository.
How to cite
If you use this data, cite the dataset, the paper that describes it, and the source corpus:
@dataset{salamon_urbansed_2017,
author = {Salamon, Justin and MacConnell, Duncan and Cartwright, Mark and Li, Peter and Bello, Juan Pablo},
title = {URBAN-SED},
year = {2017},
publisher = {Zenodo},
doi = {10.5281/zenodo.1324404}
}
@inproceedings{salamon2014urbansound,
author = {Salamon, Justin and Jacoby, Christopher and Bello, Juan Pablo},
title = {A Dataset and Taxonomy for Urban Sound Research},
booktitle = {Proceedings of the 22nd ACM International Conference on Multimedia},
year = {2014}
}
@inproceedings{salamon2017scaper,
author = {Salamon, Justin and MacConnell, Duncan and Cartwright, Mark and Li, Peter and Bello, Juan Pablo},
title = {Scaper: A Library for Soundscape Synthesis and Augmentation},
booktitle = {IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA)},
year = {2017},
address = {New Paltz, NY, USA}
}
Usage
from huggingface_hub import hf_hub_download
path = hf_hub_download("LakoreAI/URBAN-SED", "URBAN-SED_v2.0.0.tar.gz", repo_type="dataset")
Some archives of this dataset contain macOS ._* resource files; delete them before globbing for
*.wav (find . -name '._*' -delete).
- Downloads last month
- 35