InterReco is the essential choice for constructing an efficient, stable, convenient speech service.
InterReco is the key engine for Voice-Commerce; it drives speech service and search business. Our customers will apply the advance speech service solution to enhance the service quality and satisfy service requirement from their client with iFLTYTEK speech recognition technology, speech synthesis technology and professional service. Users of InterReco will experience an efficient, stable, convenient service from anywhere, anytime.
The InterReco system will perform well even with huge vocabulary database, and it will supply the confidence score of the results.

◆ Dynamic grammar

Dynamic grammar is complicated by the application program when running.

◆ VAD (Voice Activity Detection)

VAD is the process of automatic checking start point and end point of input audio. It will economize the CPU and network usage.

◆ Barge-in function

There is usually a voice prompt to indicate that the recognition engine is ready to receive a command, and people should say a command after hearing the voice prompt. However, with barge-in function, users can say a command without waiting until the voice prompt finished, which is more convenient for users.

◆ N-best results suggestion

For some applications, not only the best recognition result, but also other possibilities are needed.  InterReco can offer a list of N-best recognition results.

◆ Grammar Weight

Users are allowed to set the grammar weight of a word or a phrase, and improve the accuracy and efficiency of the system by increasing the grammar weight of keywords.

◆ Noise Reduction

InterReco can filter the background noise with its noise reduction configuration.

◆ Intelligent adjustment of recognition Strategy

The InterReco will adjust recognition strategy based on the system occupation: when system is busy, it will apply the algorithm with smaller CPU usage to guarantee the response time, when system is not busy it will choose the algorithm with high accuracy.

◆ Voice recording

It is possible for users to add command or vocabulary into dynamic grammar. It will increase the accuracy of recognition, and convenience of the user’s input.

◆ Log function

The log function presents an important role in a speech technology system. This database (Call-log) is the foundation of optimization。

◆ Speaker self adaptation

The recognition engine will adapt the accent of user to enhance the recognition rate.

◆ Adaptation Modeling

The system optimizes common model to adapt with more applications and environment.

◆ Support DTMF recognition

It will cooperate with the grammar work, recognition system will recognize the DTMF signal, and returns the recognition result.

◆ High recognition accuracy
◆ Intelligent noise reduction and voice barge-in
◆ Stability meets the requirement of Telecommunication industry
◆ Comprehensive features
◆ Complete application solution
◆ Efficient professional speech technology service