PCVC Speech Dataset

• Comment: The dataset may very well meet our notability requirements, but the article will need references to multiple, reliable, independent sources in order to show that before w can accept.
  It is often the case, in fact, nearly universally the case, that new editors find the specifics of what we need there tough to understand, you may wish to consider asking for assistance at the Wikipedia Teahouse. Thanks. joe decker^talk 16:12, 30 March 2018 (UTC)

---

The PCVC Speech Dataset is a Modern Persian speech corpus for speech recognition. The dataset contains sound samples of Modern Persian combination of vowel and consonant phonemes from different speakers. Every sound sample contains just one consonant and one vowel So it is somehow labeled in phoneme level. This dataset contains of 23 Persian consonants and 6 vowels. The sound samples are all possible combinations of vowels and consonants (138 samples for each speaker). The sample rate of all speech samples is 48000 which means there are 48000 sound samples in every 1 second. Every sound sample is 276 seconds(138 two seconds samples). In each 2s sample, in average, 0.5 second of each sample is speech and the rest is silence. In each sound sample 0.25s of start and 0.25s of end of it is surely scilence. Also in each 2s first consonant phoneme pronounced and then vowel is. All of sound samples are denoised with "Adaptive noise reduction" algorithm.

Contents

The corpus is downloadable from its GitHub web page, and contains the following:

• .mat data files of sound samples in a 23*6*30000 matrix, in which 23 is number of consonants, 6 is the number of vowels and 30000 is the length of 2s sound sample.

See also

• Comparison of datasets in machine learning

External links

• The GitHub page of PCVC speech dataset

Category:Corpora Category:Datasets in machine learning Category:Persian language