MTEB logo MTEB
classification audio

LibriCount

Cite this task

@inproceedings{Stoter_2018,
  author = {Stoter, Fabian-Robert and Chakrabarty, Soumitro and Edler, Bernd and Habets, Emanuel A. P.},
  booktitle = {2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
  doi = {10.1109/icassp.2018.8462159},
  month = apr,
  pages = {436-440},
  publisher = {IEEE},
  title = {Classification vs. Regression in Supervised Learning for Single Channel Speaker Count Estimation},
  url = {http://dx.doi.org/10.1109/ICASSP.2018.8462159},
  year = {2018},
}

Multiclass speaker count identification. Dataset contains audio recordings with between 0 to 10 speakers.

Domains: Speech
Languages: English
Task type
AudioClassification
Main metric
accuracy
Source dataset
mteb/libricount
License
cc-by-4.0
Dates
2017-01-01 → 2017-12-31
Annotations
algorithmic
Sample creation
created
Models scored …
Dataset preview mteb/libricount Open on HuggingFace →

Model scores

Loading…