InfoTalk-Speaker

World's best Cantonese text-to-speech

  • Text-to-Speech
  • TTS
  • Speech Synthesis
Overview

As a winner of multiple awards, InfoTalk-Speaker is widely accepted as the premier solution for applications that require multilingual and mixed-lingual text-to-speech (TTS). The technology for multi-cultural environments such as trilingual Hong Kong, where Cantonese, Putonghua Chinese and English are commonly spoken.

Outstanding humanlike technology

Most accurate and natural pronunciations, approaching human performance. Highest-quality multiple and mixed languages, using artificial intelligence, machine learning and deep learning technology.

Privacy & flexibility

On premises and/or cloud. Out-of-the-box and/or customisation. Impeccably integrated with the rest of the InfoTalk-RSVP family.

Why it is hard

Cantonese carries meaning in pitch.

Six tones, one syllable. Getting the contour wrong changes the word, so text-to-speech has to place every one of them correctly.
  • 詩si1
    High level
  • 史si2
    High rising
  • 試si3
    Mid level
  • 時si4
    Low falling
  • 市si5
    Low rising
  • 是si6
    Low level
Major features

A voice customers cannot tell from a person.

LanguagesCantonese, English, Putonghua Chinese. World's best Cantonese text-to-speech (TTS).
Industry standardRapid deployment and compliance. Phone networks and data networks. Customers across many industries.
Performance & scalabilityScalable operation. Load-balancing architecture. Disaster recovery.
Sibling voice productsInfoTalk-RSVP, InfoTalk-Recognizer, InfoTalk-Vbrowser, InfoTalk-Processor
Applications

Ideal for these applications.

  • Call centres
  • Chatbots
  • Voicebots
  • Talking robots
  • Customer services
  • Intelligent IVR systems
  • Automated receptionists
  • Transcription
  • Text-readers
  • Audio books
  • Voice devices
  • Talking kiosks
Seven facts

7 facts about Cantonese and Chinese text-to-speech

What is text-to-speech?

Text-to-speech is about converting texts into a human voice. You type the texts, and the technology converts your texts into a human voice.

What is multilingual text-to-speech?

In a multi-cultural multilingual environment like Hong Kong, people often mix English words with Cantonese or Chinese words when they speak. Some English words are so common in daily conversations that even native speakers would not remember the Cantonese or Chinese equivalent. Even if the text contains both Chinese and English characters, multilingual text-to-speech speaks both languages in the same natural human voice.

Why is text-to-speech challenging?

Text-to-speech has been commercialised by MIT technologists since the 1980s. However, the voice has been robotic, making it suitable for limited and special applications only. More recently, great strides have been made in artificial intelligence and machine learning. The combination of such advances and decades of speech knowledge completely revolutionises the text-to-speech industry. Today, you can hardly hear the difference between the voices of the human and text-to-speech.

Why is text-to-speech in Cantonese and Chinese more challenging than in English?

As a tonal language, Chinese adds an extra layer of hurdle for speech technologists to overcome. Being more like a verbal language than a written language, Cantonese raises the bar even higher. The complex interaction between its intonation, idiosyncrasy and colloquialism makes the development of Cantonese text-to-speech difficult for many technologists. However, committed to Hong Kong, InfoTalk has developed the world's best Cantonese text-to-speech.

What is the difference between artificial intelligence (AI) and text-to-speech?

Text-to-speech has always been a multi-disciplinary discipline involving computer science, engineering, signal processing, acoustics, language and linguistics, among others. With recent technology advances in artificial intelligence (AI), machine learning (ML) and deep neural networks (DNN), text-to-speech has become more interdisciplinary than ever before.

Is text-to-speech still in the laboratory stage?

Text-to-speech technology in Cantonese, Putonghua and English has been successfully deployed in a wide variety of systems and industries in Hong Kong and Asia. Successful commercial deployment, especially in mission-critical high-traffic applications for public use, requires experience, expertise and knowhow that are above and beyond simplistic laboratory experiments.

Does text-to-speech have other names?

There are other names for text-to-speech (TTS), such as text-to-voice, speech synthesis and voice synthesis. Chinese and Cantonese names in Hong Kong, China and Asia include 文語轉換, 文字轉語音, 語音合成. They all mean more or less the same technology. However, Hong Kong text-to-speech is unique in that people speaking Cantonese often mix it with English, so it speaks Cantonese mixed with English. The same applies to Putonghua text-to-speech.

Get in touch

See it working on your own material.

Bring one workflow and a sample of your recordings. We strive to respond within 24 hours.

OFFICESHong KongNorth ChinaSouth ChinaTaiwanSoutheast AsiaTEL (852) 6300 6988