Nine BUT phone recognizes?

[Ahmed Ali]

Hi all,

I was reading this paper  "SUPERVECTOR PRE-PROCESSING FOR PRSVM-BASED CHINESE AND ARABIC DIALECT IDENTIFICATION" 

And they cites this package and mentioned that  PRSVMs based on nine BUT phone recognizes...

I can see only here 4 systems here:

• PHN_CZ_SPDAT_LCRC_N1500 - 8kHz, 2 block STC, trained on Czech SpeechDat-E, 15 banks, 31 points, the DCT is applied on each temporal vector to reduce its size to 11 values, 1500 neurons in all nets
• PHN_HU_SPDAT_LCRC_N1500 - 8kHz, 2 block STC, trained on Hungarian SpeechDat-E, 15 banks, 31 points, the DCT is applied on each temporal vector to reduce its size to 11 values, 1500 neurons in all nets
• PHN_RU_SPDAT_LCRC_N1500 - 8kHz, 2 block STC, trained on Russian SpeechDat-E, 15 banks, 31 points, the DCT is applied on each temporal vector to reduce its size to 11 values, 1500 neurons in all nets
• PHN_EN_TIMIT_LCRC_N500 - 16kHz, 2 block STC, trained on TIMIT, 15 banks, 31 points, the DCT is applied on each temporal vector to reduce its size to 11 values, 500 neurons in all nets

Does any one know where are the other five recognizes/systems?

Thanks,

Ahmed