Saturday , 23 February 2019

New IBM Text To Speech Features Engage Users With Expressive Content

Text to speech technology is far from new. However, there have been recent developments in this area that have brought renewed attention to the technology. If you are a developer, you have definitely already heard about IBM Watson. But, did you know that Watson AI technology led to the creation of an all new IBM text to speech service? Learn all about the new IBM Watson text to speech service features and how it works below.

What Is IBM Text To Speech?

IBM’s new Text to Speech program makes it easier than ever for web and app developers to generate synthesized audio output. This technology is far from new or revolutionary. The incredible thing about IBM’s new Text to Speech service is that it actually produces synthesized audio that uses the right intonation and and cadence. Simply put, it helps produce audio output that actually sounds the way a normal person would speak. The audio produced by the IBM service is streamed to listeners right as it is being produced. There is virtually no delay. Put simply, the Watson Text to Speech platform makes it possible to have your systems speak to users in a more audibly-pleasing manner. This technology opens up a whole new world of possibilities for developers like you, just as radio broadcasting did in the past.

Audio Types

The synthesized audio produced by IBM Text to Speech is only available in a small number of audio formats. You can use the service to produce audio in Ogg or WebM using the Opus or Vorbis codec. In addition, you can produce WAV, FLAC, MP3 (MPEG), l16 (PCM) or mulaw as well. Of course, basic format audio can also be the chosen format for your synthesized audio output. Be sure that your own programs you are developing are compatible with at least one of these audio formats. That way, you can take advantage of the incredible IBM Text to Speech platform.

Supported Languages

Of course, IBM’s Text to Speech technology is only offered in certain languages as well. Compatible languages include French, German, Italian, Portuguese and Japanese. Other supported languages include both American English and UK English. The service even includes two different options for Spanish, Castilian Spanish and North American Spanish. Surely as the technology progresses in the coming months and years, even more languages will be offered. That will help developers like you reach even more users worldwide with your mobile apps and software.

Expression

Some Text to Speech languages even offer the option to select an emotional expression for the audio you create. These expressive language functions allow you to select an emotion you wish for the AI voice to convey to listeners. This expressive SSML, or Speech Synthesis Markup Language, is limited, however. You can select from three different emotions appropriately titled GoodNews, Apology or Uncertainty. These emotional expressions could theoretically make it much more exciting to have your AI personal assistant read your media mentions alerts to you. This is the truly revolutionary aspect of the IBM SSML utilized in the Watson Text to Speech platform.

Customization

IBM Watson Text to Speech makes it possible for SSML develops to customize pronunciations. This is incredibly important for those programs that use their own unique vocabulary. Developers can use the platform to customize pronunciations down to the individual syllable. The ability to customize word pronunciations with technology gives you entire control over every single aspect of your software. Never before has there been so much flexibility when it comes to AI voices and other types of text to speech audio output. If you want to manage every little detail on your own, Watson’s Text to Speech service offers the customization capabilities that such control requires.

If you are a software developer or any other type of developer, you may benefit from trying out the new IBM Text to Speech technology. The new IBM audio production tool makes it easier than ever to provide a personalized experience to program users. Thanks to the new Expressive SSML technology developed by IBM, you can now produce audio outputs that engage app users almost as well as a real person. Learn up on the service’s features mentioned above before you get started developing in SSML. It could definitely prove useful if you develop Internet of Things apps in the future.

Photo from https://www.ibm.com/blogs/watson/2016/09/build-custom-language-model-convert-unique-speech-text/

Leave a Reply

Your email address will not be published. Required fields are marked *

*

Scroll To Top