Our websites:

Our social media sites - YouTube, Twitter and Facebook

YouTube X Facebook Instagram
2068

Vibes of AAC: Adding Expression and Emotion to AAC Voices

Posted by Ceanna MacGregor on the 28th October, 2024

Category AAC Communication Support Needs

It's not just what you say, it's how you say it. Until recently, options were limited for adding rich intonation to text-to-speech (TTS) voices. As the theme for AAC Awareness Month is 'Vibes of AAC,' it seems fitting to explore some ways to add a 'vibe' when communicating with AAC apps and devices.

Voice and Identity

Our voice is an important part of our identity. When selecting a voice to use with an AAC device, it is important to consider aspects such as accents, age, pitch and rate, with the AAC user's views and preferences at the forefront of decisions. 

As well as expressing thoughts and emotions, our voice is also important in conveying our personality. An excellent example of this is with Christian Wilson who has recently received his dream voice. He selected two voice donors and engineers at Speak Unique blended their voices to create the accent he was searching for.

Read more about Christian's story.

We communicate so much more than just the words we say with how we say them. For example, our non-verbal communication such as facial expression and body language adds to the emotion of a message. The way we say something can completely change the meaning of the message or how it is received. 
 
Let's use the following example: "That's great." Think of how it might sound if you are:

It's likely the tone and inflection vary. (And probably facial expression and body language too!)
 
A rise in intonation at the end of a phrase can highlight the difference between a declarative statement and a question.
 
For example: 


      

Adding Expression to Text-to-Speech Voices

 

So, how can we provide the same opportunities for those who speak with communication apps and devices? As technology continually advances, so do the quality and selection of voices available to those who use communication devices and screen readers. 
 
Some options are the CereProc and Acapela Group voices which provide high quality voices in various languages. CALL Scotland, in collaboration with CereProc and with funding from the Scottish Government, provide Scottish digital voices to use with Windows devices for free for the Scottish public sector. Find out more about Scottish digital voices on The Scottish Voice website.

Both CereProc and Acapela have additional ways to include natural sounds in speech.

CereProc have 'voice gestures' that work with Scottish voices on Windows devices. These are a collection of sounds and short words we say when speaking. There are specific 'tags' for each sound that when added to specific text, generate the related sound.

Acapela have 'Voice Smileys' that work on both iOS and Windows devices for many of their voices. There are 'exclamations' which are words and short phrases and 'sounds.' Exclamations require only an exclamation or question mark added to specific words or phrases. Acapela sounds work in a similar way to CereProc tags and consist of some text between hash signs. 

Check out some examples on our Expression in AAC YouTube playlist.

      

In recent blog posts, members of the AssistiveWare team have shared how you can add expression to some Acapela voices from within their AAC apps.


AI and AAC

The power of AI is already transforming text-to-speech (TTS) and personal voice options as well as features in AAC apps and software. The possibility of using AI to enhance the speed of creating messages while adding personality is currently being explored within the AAC community.  One example is the following study:

“The less I type, the better”: How AI Language Models can Enhance or Impede Communication for AAC Users.

Authors: Stephanie Valencia, Richard Cave, Krystal Kallarackal, Katie Seaver, Michael Terry, Shaun Kane. 

CHI ’23, April 23–28, 2023, Hamburg, Germany © 2023 Copyright held by the owner/author(s). ACM ISBN 978-1-4503-9421-5/23/04. https://doi.org/10.1145/3544548.3581560


In their study, the authors consider some benefits, challenges and considerations of the use of AI technologies, such as Large Language Models, for AAC users. From their abstract, they say:

Our study participants believed that AI-generated phrases could save time, physical and cognitive effort when communicating, but felt it was important that these phrases reflect their own communication style and preferences. This work identifies opportunities and challenges for future AI-enhanced AAC devices. 

Watch a presentation summarising their work.

Access the full research article.

 

At AAC in the Cloud 2024, Avaz AAC shared some interesting insights into some possibilities of integrating AAC and AI: Creating an Authentic & Efficient AAC system powered by Artificial Intelligence.

They have since released their 'Expressive Tones' feature, although not yet available with Scottish voices, they are certainly an exciting prospect for AAC users. You can see an example here: 

Find out more about Expressive Tones on Avaz AAC.

With continual advancements in technology, the potential for more control over what you sound like and how you say what you want to say as an AAC user seems likely to evolve and improve.

Technology to Support Dyslexia in Education

5-week short study online course

Newsletter: join thousands of other people

Once a month we'll send you an email with news, research and thoughts, as well as training courses and free webinars you may wish to attend.