ecur 得意音通技术 16 Data Creation- Read Speech (4) Your Partnerinthe Century of speech 二二 Average -Natual --RH Average -Natural --RHF- ELF n this example, all seen di-IFs are sorted in an ascending order of their occurring counts in the mother set. The occurring count as a function of the di-IF order is shown in the above Figure: (a), the first 90% di-IFs, and (b), the next 10%di-IFs
Your Partner in the Century of Speech 16 Data Creation – Read Speech (4) 0 100 200 300 400 500 600 Average Natual RHF ELF 0 500 1000 1500 2000 2500 3000 3500 4000 4500 Average Natural RHF ELF (a) (b) In this example, all seen di-IFs are sorted in an ascending order of their occurring counts in the mother set. The occurring count as a function of the di-IF order is shown in the above Figure: (a), the first 90% di-IFs, and (b), the next 10% di-IFs
ecur 得意音通技术 17 Data Creation- Read Speech 5) Your Partnerin the Century of Speech 口RHF: f helpful to make distribution as close to the average curve as b possible s It cannot guarantee a minimal occurring count for those less frequently seen di-IFs 口ELF: i guarantee a minimal occurring count so that data is enough to train the models for those less frequently seen units Will lead to relatively big occurring counts for those most frequently seen di-IFs though the distribution is close to natural uIt is expected that a good result can be achieved by combining both the rhf and the elF strategies
Your Partner in the Century of Speech 17 Data Creation – Read Speech (5) ❑ RHF: ❖ helpful to make distribution as close to the average curve as possible ❖ It cannot guarantee a minimal occurring count for those less frequently seen di-IFs. ❑ ELF: ❖ Guarantee a minimal occurring count so that data is enough to train the models for those less frequently seen units ❖ Will lead to relatively big occurring counts for those most frequently seen di-IFs though the distribution is close to natural. ❑ It is expected that a good result can be achieved by combining both the RHF and the ELF strategies
ecur 得意音通技术 18 Data Creation- Read Speech (6) Your Partnerin the Century of Speech aQuality control is very important, for read speech database, there could be three steps ☆Self- check ☆ Cross- check ☆ Final- review
Your Partner in the Century of Speech 18 Data Creation – Read Speech (6) ❑Quality control is very important, for read speech database, there could be three steps: ❖Self-check ❖Cross-check ❖Final-review
ecur 得意音通技术 19 Data Creation-Spontaneous/Conversational Speech(1) Your Partnerin the Century of Speech OIn spontaneous(casual) speech or conversational (dialogue) speech 8 More spontaneous speech phenomena (phonetically and acoustically B Richer spoken language phenomena(linguistically aMan- Man dialogue different from Man-Machine dialogue
Your Partner in the Century of Speech 19 Data Creation – Spontaneous/Conversational Speech (1) ❑In spontaneous (casual) speech or conversational (dialogue) speech ❖More spontaneous speech phenomena (phonetically and acoustically) ❖Richer spoken language phenomena (linguistically) ❑Man-Man dialogue different from Man-Machine dialogue
ecur 得意音通技术 20 Data Creation-Spontaneous/Conversational Speech(2) Your Partnerin the Century of Speech u Three kinds of scenarios should be considered Spontaneous(Casual Giving speech or talking about something Man-Machine conversation conversation mode, but it is important in telephony speec true Leaving a voice message: leaving a voice message is not a Man-Machine dialogue: given a simulated service provider, the speaker can inquire about whatever he/ she wants in the following domains Hotel/ticket reservations Telephone number inquiries Stock prices inquiries Weather inquiries City route map inquiries Telephone-based shopping/ordering
Your Partner in the Century of Speech 20 Data Creation – Spontaneous/Conversational Speech (2) ❑ Three kinds of scenarios should be considered ❖ Spontaneous (Casual): ▪ Giving speech or talking about something ❖ Man-Machine conversation: ▪ Leaving a voice message: leaving a voice message is not a true conversation mode, but it is important in telephony speech. ▪ Man-Machine dialogue: given a simulated service provider, the speaker can inquire about whatever he/she wants in the following domains: – Hotel/ticket reservations – Telephone number inquiries – Stock prices inquiries – Weather inquiries – City route map inquiries – Telephone-based shopping/ordering –