Project description: Embedded Speech Corpus

---

◦ Speech collected through USB-DSP collection toolkit for developing 
speech recognition application in embedded environment
◦ A total of 352 speakers:
◦ for training: 300 speakers, for training: 52 speakers
◦ Office
◦ Microphone: CMT-58 ECM mic
◦ Sampling and data format
 - 16,000 Hz, 16 bit Windows wave format
◦ Prompts
 - 12 single digits
 - 12 PDA commands (for all the speakers) 
 - 4,175 phonetically rich words
◦ Prompts Sheet
 - single digits, PDA commands 
 - 50 sets of phonetically rich words
◦ Per speaker
 - single digits, PDA commands
 - 1 set of 107-108 phonetically rich words