Datasets:

Matthijs

/

cmu-arctic-xvectors

like 37

In the dataset, each record has a filename like cmu_us_awb_arctic-wav-arctic_a0001.

My interpretation of this is:

• cmu : Carnegie Mellon University, which created the dataset
• us : The type of English is US English. Surprisingly, even the Scottish English male speaker is described as using US English!
• awb : One of seven codes indicating the category of speaker. The seven categories, listed on the model card, are: bdl (US male); slt (US female); jmk (Canadian male); awb (Scottish male); rms (US male); clb (US female); ksp (Indian male). The counts for each are

Speaker origin                                         
Canadian male         1132 
Indian male           1132 
Scottish male         1138  
US female             2264  
US male               2265 

• arctic-wav-arctic : it is the ARCTIC dataset and wave files. Unclear why arctic is repeated
• a0001 : reference to the particular speaker. One character, a or b, followed by 4 digits.

This pattern holds true for all of the items in the dataset, apart from #2270 with filename cmu_us_bdl_arctic-wav-hf_gen. Instead of an ending like arctic_a0001, it has hf_gen. Can anyone shed light on this?