◆ Convenient and flexible application architecture
Efficient network speech synthesis service and centralized resource management mechanism based on TCP/IP form an organic combination framework of client, resource manager and server, and construct a flexible and extensible scheme. Its high availability has been verified in large-scale key business applications in several key industries, ensuring reliable 7×24-hour uninterrupted automatic voice service. It also supports distributed application architecture (patented technology). The front-end responsible for text analysis and preprocessing and the back-end responsible for speech synthesis can be deployed on remote servers, and only the analyzed and labeled text information can be transmitted between remote networks, which greatly reduces the network bandwidth requirements of voice applications and is very suitable for large-scale distributed voice applications based on the Internet. Flexible and efficient development interface
According to different development tools, different integration requirements and schemes, Interactive SDK provides a variety of development interfaces, including standard development interface (DLL), COM component, SAPI development interface and so on. Developers can choose flexibly according to actual needs. Provide rich development routines and documents to help partners accelerate the development process of voice applications.
◆ Rich parameter setting and flexible adjustment function.
Provide rich and perfect dynamic parameter setting and adjustment functions and tools to help users flexibly and efficiently control and manage the speech synthesis effect. Provide tools for unified configuration and management of global parameters (such as volume, speech speed, pitch, etc.). ), user dictionaries, user rules and customized resource packages; Setting of numbers, punctuation marks and English pronunciation; The function of adding Chinese and English words can specify the pinyin or phonetic symbols of each word. Provides a unified and easy-to-use graphical user interface for operation settings, which can be dynamically set and adjusted through API parameters, and also supports CSSML (Chinese Speech Synthesis Markup Language) for marking, description and control.
◆ Support open standards
Fully support the General Technical Standard for Chinese Speech Synthesis System (GB/T2 1024-2007), and follow the definition of terms, classification standards, data exchange format standards and application specifications stipulated in the standard.
SSML is a part of W3C's voice interface framework, which is a set of specifications about voice applications and building voice applications on the World Wide Web. Through SSML, people can listen to synthesized speech more through mobile phones, desktop computers and other devices, and extend computing and information transmission to every corner of the world.
It supports Media Resource Control Protocol (MRCP), which is published by IETF and defines the interface standard between media servers and network voice resources (including voice recognition and speech synthesis servers).
◆ Efficient and convenient enhanced tool set
Iflytek has accumulated rich practical experience in the process of helping customers develop applications and optimize effects for a long time. On this basis, a series of convenient and efficient components have been gradually formed, such as offline voice application tools, CSSML visual editing tools, DOC/XLS text format conversion tools and so on. Flexible use of these tools is helpful to accelerate application development, optimize synthesis effect, and facilitate system maintenance and technical support.
◆ Character set and voice data format support
Fully support GB23 12, GBK, BIG5, GB 18030, UTF-8 and UNICODE coded character sets, and automatically recognize UNICODE texts; Support the direct output of voice data in various formats (including 6k/8k/11k/16k), such as linear Wav, A/U Wav, Vox, etc.
◆ Wide platform support
Support mainstream operating systems, the server supports Windows, Unix, Linux and other operating systems, and the client supports Microsoft Windows, SUN Solaris, REDHAT Linux, SUSE Linux and other operating systems.
There have been successful integration cases with well-known related platforms and equipment vendors in the industry. Through close cooperation with many platform and equipment providers, system integrators and software developers, we can guarantee to provide users with professional services around the whole process of voice application.
◆ CSSML, the effect can be improved more freely.
Cssml (Chinese Speech Synthesis Markup Language) is a Chinese speech data description specification proposed and led by Iflytek. The standard has been highly valued and supported by the National 863 Expert Group, the State Information Commission and the State Bureau of Technical Supervision. In 2005, it officially passed the evaluation of the National Organization for Standardization and became an important part of the technical standards and norms of Chinese speech synthesis. CSSML is designed and extended for the application of Chinese Pinyin, which can flexibly mark and control various features and is compatible with SSML.
Pre-recorded voice, smooth connection and simple application.
InterPhonic provides innovative unified management function of pre-recording, which takes pre-recording as the resource of speech synthesis system. Through intelligent matching of prompt tone and synthesis template, the matching between pre-recording and synthesis speech is simpler and smoother, and at the same time, it avoids frequent switching and transition between prompt tone playback and speech synthesis, simplifies the complexity of application process, and further improves the service effect and quality.
◆ Background music quickly improves the user's physical examination.
InterPhonic provides the industry's first background sound function. Through the simple and easy-to-use tools provided by the system, you can add background music conveniently and efficiently, adjust the volume contrast between background music and synthetic voice, and directly listen to the actual effect, making the voice service more friendly and natural.