CN EN
CN EN
Core Techniques

Automated Speech Recognition

The Automated Speech Recognition(ASR) technologyof TalentedSoft adopts the most advanced deep learning algorithm within the industry, and hasdeveloped thenew-generation ASRindependently.Specifically,the Products can be customized according to different scenarios. TalentedSoft provides a one-stop process and assists the cooperation partners to achieve professional development in various domains.

Voiceprint Recognition

For the VoicePrint Recognition (VPR) technology,TalentedSoft adopts the core Algorithms accumulated for more than ten years . It has independently developed the text-independent voiceprint recognition with short utterances, text-dependent voiceprint recognition with short utterances and voiceprint recognition technology with digital dynamic password. In order to achieve fast and accurate identity verification, TalentedSoft provides different kinds of Speaker Verification (1:1) and Speaker Identification(1:N) services, which is widely used in remote identity verification. At the same time, TalentedSoft also provides the technology of Speaker Diarization, which can be applied for the role separation of conference recording and customer service calls.

Text to Speech

The Text to Speech (TTS) technology of TalentedSoft is based on the continuously optimized end-to-end speech synthesis algorithm, which can convert the input text information into smooth and natural speech. TalentedSoft has launched the suitable solutions for different fields, industries and scenarios.

Language Identification

The Language Identification (LID) technology of TalentedSoft is based on the advanced deep learning algorithms within the industry. And TalentedSoft has developed the multilingual recognition engine . Currently, languages available to be recognized include: Chinese, English, Hokkien, Cantonese, Tibetan, Uyghur, Kazakh, Japanese, Korean, Russian, Vietnamese, Indonesian, etc.

Voice Wake-up

The Voice Wake-up technology of TalentedSoft presets the wake-up word on the smart device.When the user issues a wake-up voice command, the device would be woken up from thesleepmode andgivea respond. Such process begins the first step of human-computer interaction.

Gender Recognition


The Gender Recognition technology of TalentedSoft classifiesthespeechbased on the statistical models, which can quickly identify the gender of the speaker in real time.

Automated Speech Recognition

The Automated Speech Recognition(ASR) technologyof TalentedSoft adopts the most advanced deep learning algorithm within the industry, and hasdeveloped thenew-generation ASRindependently.Specifically,the Products can be customized according to different scenarios. TalentedSoft provides a one-stop process and assists the cooperation partners to achieve professional development in various domains.

Voiceprint Recognition

For the VoicePrint Recognition (VPR) technology,TalentedSoft adopts the core Algorithms accumulated for more than ten years . It has independently developed the text-independent voiceprint recognition with short utterances, text-dependent voiceprint recognition with short utterances and voiceprint recognition technology with digital dynamic password. In order to achieve fast and accurate identity verification, TalentedSoft provides different kinds of Speaker Verification (1:1) and Speaker Identification(1:N) services, which is widely used in remote identity verification. At the same time, TalentedSoft also provides the technology of Speaker Diarization, which can be applied for the role separation of conference recording and customer service calls.

Text to Speech

The Text to Speech (TTS) technology of TalentedSoft is based on the continuously optimized end-to-end speech synthesis algorithm, which can convert the input text information into smooth and natural speech. TalentedSoft has launched the suitable solutions for different fields, industries and scenarios.

Language Identification

The Language Identification (LID) technology of TalentedSoft is based on the advanced deep learning algorithms within the industry. And TalentedSoft has developed the multilingual recognition engine . Currently, languages available to be recognized include: Chinese, English, Hokkien, Cantonese, Tibetan, Uyghur, Kazakh, Japanese, Korean, Russian, Vietnamese, Indonesian, etc.

Voice Wake-up

The Voice Wake-up technology of TalentedSoft presets the wake-up word on the smart device.When the user issues a wake-up voice command, the device would be woken up from thesleepmode andgivea respond. Such process begins the first step of human-computer interaction.

Gender Recognition


The Gender Recognition technology of TalentedSoft classifiesthespeechbased on the statistical models, which can quickly identify the gender of the speaker in real time.

Solutions

Intelligent Customer Service

The intelligent speech customer service system combined with technologies of TalentedSoft like Automatic Speech Recognition (ASR), Text to Speech (TTS) Voiceprint Recognition (VPR) etc., which has a strong capability of vertical field customized service and can meet the growing service requirements of information consultation, telephone transactions,etc.

Intelligent Conference

The intelligent speech conference system can translate voice into text in real time. The core technology of the intelligent speech conference system has reached the commercial level. It is suitable for various conference scenarios and can help generate the meeting summary quickly and efficiently.

Identity verification

TalentedSoft can use the Voiceprint Recognition technology, combined with specific application scenarios to flexibly choose the methods of identity verification, such as text-independent or dynamic password, mobile terminal or telephone and other authentication methods.

Audio reading

The technology of Text to Speech could help develop the audio reading industry. Literary works, news, information and other text content could be read emotionally. The reading voice can also be customized.
Advantages

Full stack self-research technology

TalentedSoft spared no effort in researching and developing speech technologies like Automated Speech Recognition, Text to Speech and Voiceprint Recognition. These technologies are widely used in various industries, and there are also a large number of application cases and speech data accumulation.

Support multi-platform access

TalentedSoft provides the REST API service, supports intelligent hardware terminals, mobile applications, website applications, desktop applications and other equipment terminals that need to be accessed.

Customization & Privatization

TalentedSoft provides in-depth customized services, it can meet the requirements of customization of particular industry business, making the services better. TalentedSoft also supports privatization and relating services, ensuring the security of customer data information.

Comprehensive technical services

TalentedSoft provides comprehensive and professional technical support, with diversified options and detailed documentation. It provides customers and developers with a full range of technical services.
About Us

Xiamen TalentedSoft Co., Ltd. (abbreviation: TalentedSoft) was established in 2005. It has been engaged in the research of speech technology over a long period, and has always kept up with the international advanced technology level. It takes an active part in promoting the development and application of artificial intelligence(AI)products. It also provides speech interaction solutions and technical services suitable for multiple industries, multidomain and multiple sceneries.


The researching and developing team of TalentedSoft is led by well-known scientists within the industry, most of whom have the research and development experience of the field of more than 15 years.The team members are mainly from Xiamen University, Chinese Academy of Sciences and other well-known research institutes. TalentedSoft and Xiamen University together set up the "TalentedSoft-XMU Speech Lab", sparing no efforts in researching and developing core technologies, and cultivating top-level talents, together promoting the research of speech technologies and industry development.


TalentedSoft independently developed and established the intelligent speech interaction platform and launched a series of artificial intelligence technologies, including Automated Speech Recognition(ASR), Voiceprint Recognition(VPR), Text to Speech(TTS), LanguageIdentification (LID),Gender Recognition, and Voice Wake-up. It has provided services for dozens of companies in total, and its products have been widely used in public security justice, social security certification, intelligent customer service, intelligent medical care, intelligent securities, intelligent dispatching and other fields.


As a technologically innovative enterprise, TalentedSoft focuses on researching and developing, and keepsupinnovating. It has achievedpromisingresults inrelatedchallenges ofVoiceprint Recognition, Language Recognition and Automated Speech Recognition in and abroad. At the same time, TalentedSoft is also the cooperating enterprise with the Ministry of Public Security inbuilding thesecurity voiceprint recognition standard.


By the efforts and accumulation for more than ten years, the outstandingachievementsof TalentedSoft in intelligent speech technology research and industrialization have been widely recognized bythe society. TalentedSoft is willing to cooperate with partners to promote artificial intelligence into thousands of households and build a better future together.