| |
|
|
| |
> |
KTV Song Selection by Speech Solution |
| |
 |
| |
|
ThinkIT song selection by speech (S3) solution functions as a song selection system for KTV entertainment, either by selecting via speech, the name of a song or a singer, or by humming the melody, or by reading lyrics of a wanted song, the solution can also function as a singing scoring system in KTV rooms. Due to its simple use and easy switch over between operations, this solution is now being applied in KTV entertainment to replace the traditional song selection system. |
| |
|
more |
| |
> |
Telephony Speech Recognition Engline |
| |
|
| |
|
It is designed specifically for telephony speech recognition. The kernel is based on Hidden Markov Model (HMM), and the search algorithm has been optimized greatly for speed and accuracy suitable to the characteristics of telephony speech. The Chinese acoustic model was obtained via training speech data with a variety of environments, speakers and accents. So the model should be robust to some extent. Our engine also has the confidence measure technology which can reject most of the unusual speech. |
| |
|
more |
| |
> |
Telephony Automatic Speech Recognition |
| |
|
| |
|
Our automated telephony system (TIDS) is a computerized and SR based telephone operator, which successfully puts our world leading SR technology into telephony applications. This system can automatically transfer phone calls to the wanted persons, with no need of human operator. When a call comes in, all what the caller needs to do is just speaking to the system the name of a wanted person, the system will then instantly recognize the wanted person and connect the extension phone through. This system is an ideal choice for those using operator-assisted phone service. |
| |
|
more |
| |
> |
Embedded Speech Recognition System |
| |
|
| |
|
The ThinkIT Embedded Speech Recognition Engine (MSR) is designed especially for mobile equipments such as PDA, cellular phone, etc. The kernel is based on Hidden Markov Model (HMM), and the search algorithm has been optimized greatly for speed and accuracy suitable to the characteristics of embedded mobile devices. The Chinese acoustic model was obtained via training speech data with a variety of environments, speakers and accents. So the model should be robust to some extent. According to resources of different mobile equipments, the recognition engine can be customized with different configurations so that the overall performance is best to certain equipments. |
| |
|
more |
| |
> |
Embedded Speech Synthesis System |
| |
|
| |
|
The ThinkIT Embedded Speech Synthesis Engine (MTTS) is designed specially for mobile devices such as cellular speech applications. It is modeled on Chinese full syllable and certain particular units. Speech compression and encoding algorithm has also been integrated. The engine has been optimized in speech database size and naturalness suitable for mobile applications with limited resources. The speech database can also be customized according to varieties of resources availability and application requirements so that the overall performance will be optimal in certain applications. |
| |
|
more |
| |
> |
voice tone |
| |
|
| |
|
Our VoiceTone is a such an engine that it combines our embedded recognition engine (MSR) and MTTS together, designed specifically for high end cell phones (such as PDA phone and smartphone), providing an easy way to operate mobile device via SR commanding. User just say a wanted name, our VoiceTone will show the relative information about the wanted, and dial the wanted's number at a voice hint. Our VoiceTone works on embedded OS with very limited resource, such as Windows CE or equivalent OS, it can complete recognition, commanding and other tasks under very limited resource. The VoiceTone is purposely designed for interface on embedded systems, it makes the operation of mobile systems even more humanized, especially makes those aged and disabled people easily and conveniently use mobile devices. |
| |
|
more |
| |
> |
speaker identification engine |
| |
|
| |
|
Our speaker identification engine (TSIE) integrated our leading speech-signal processing technology with statistic modeling technology, it's capable to identify a speaker just by analyzing a short voice message. This engine can meet the special needs purposed for web safety, state security and speaker identification, etc.. |
 |
| |
|
| > |
Speech Morphing Solution |
 |
| |
Our Morphing Solution enables speaker to raise or lower voice tone at his/her own choice, it's mainly a feature for fun and lets speaker enjoy the funny effect of his/her own voice changing. |
| |
more |
| |
|
|
