ASR Customization

ASR Customization (ASRC) leverages deep learning technology to optimize speech recognition for specific fields and allows you to define language models. In addition, this service can be customized based on the requirements of a specific enterprise to help the enterprise build unique competitiveness and improve human-machine interaction experience.

Fees start as low as ¥0.004 per API call.

Learn more
Product Advantages
  • High Recognition Rate

    ASRC utilizes the deep learning technology to optimize the corpus for specific scenarios and fields, enabling an industry-leading recognition rate.

  • Cutting-Edge Technologies

    ASRC combines mature speech recognition algorithms currently in active use in the industry with the latest research to empower enterprises with unique competitive advantages.

  • Hot Word Recognition

    You can upload wordlists critical to your industry so that your professional jargon can be more accurately recognized.

  • Customizable Models

    ASRC increases accuracy by using speech recognition models designed for the specific requirements of the vertical industry you operate in of for other specific scenarios.

Application Scenarios
  • Voice Customer Service Inspection

  • Meeting Minutes Taking

  • Voice Message

  • Gaming and Entertainment

Voice Customer Service Inspection

Voice Customer Service Inspection

ASRC recognizes speech from customer service personnel and from customers, converts it into text, and checks whether the speech contains any improper content, inappropriate words, or phone numbers.

Advantages

Accurate

High recognition accuracy

Speaker Separation

Speaker separation technology enabling higher inspection efficiency

Meeting Minutes Taking

Meeting Minutes Taking

ASRC quickly recognizes the audio of meeting minutes and converts it into text, facilitating automatic and efficient meeting minutes recording.

Advantages

Multiple Input Sources

Transcription of mono- or dual-channel audio with an 8 kHz or 16 kHz sampling rate

Efficient and Convenient

Meeting minutes recorded quickly and easily

Voice Message

Voice Message

ASRC converts voice messages you send or receive into text to deliver higher reading efficiency and interaction experience.

Advantages

Accurate

High recognition accuracy

Multiple Audio Formats

Nine audio formats are supported, such as WAV, AMR, and AMR-WB.

Gaming and Entertainment

Gaming and Entertainment

ASRC converts voice chats into text messages, improving reading efficiency and user experience.

Advantages

Accurate

High recognition accuracy

Hot Word Recognition

Uploading of gaming wordlists for higher recognition accuracy

Usage Guides

Create an Account and Experience HUAWEI CLOUD for Free

Register Now