FAQ

Our primary data collection program is focused on English, but we have started small pilot programs in French, Hindi, Japanese, and Spanish. Please fill out the relevant form if you’re interested in contributing in those languages:

• French

• Hindi

• Japanese

• Spanish

Unfortunately, we cannot accept speech samples from anyone under the age of 18. However, we’ve heard incredibly strong stories highlighting how children and young adults would benefit from being a part of Project Euphonia, so we’re actively exploring how to open up our research to people of all ages.

Yes. We welcome and appreciate whatever time you’re able to contribute toward our research.

Most of the questions asked in the screener form after you click on the blue “Record phrases” button are optional, though we do require your email address (so we can contact you) and the country and state (US) you reside in (for legal reasons). These details also enable us to give you rewards for recording phrases.

We need to be able to contact you directly to troubleshoot any issues, and we use your email address to send you rewards. You can ask us to remove your email address from our records and disassociate your voice recordings from the name you provide in the sign-up flow.

Only Google employees and affiliates working on Project Euphonia will have access to your recordings. Your recordings will not be available to anyone else unless you give us explicit permission.

Yes, participants can request a copy of their own data. Please contact us with this request and allow some time for verification and processing.

Yes. Should participants want to delete the voice recording they provided, you can contact us with your request. Please review the Google Privacy Policy for more information. 

Everyone is encouraged to go at their own pace, but we estimate that recording 300 phrases without any breaks may take between 30 to 90 minutes. However, you do not need to record these in one sitting. All phrases will be saved, and you can always pick up where you left off. You can take as much time as needed to complete recording the full phrase set.

From the learnings of our research, we have created an Android app designed to help those with atypical speech communicate better. Find out more at g.co/ProjectRelate.

You can learn more about the results of this research in our publications section.

• Internet Connection

• A mobile phone OR laptop/desktop computer with a built-in or external microphone

Please watch this 2 minute video for tips on how to record phrases. We carefully listen to the initial phrase set and provide feedback to participants on recordings as needed (e.g., “Please record in a quieter room” or “Please place your phone closer to your mouth”).

Our phrases are focused on specific use cases, such as when you are using an assistive tool like Google Home or Google Assistant, or having a conversation with a loved one. All the words are carefully selected to propel our research forward.

Try pressing the “Record Phrases” button at the upper right in a Chrome Incognito Window or Safari Private Browser. If you're still having trouble, please contact us.

Yes, you can still participate if someone helps you navigate the interface of our recording tool, or if you have indirect access (e.g., via head mouse or eye gaze) to a computer with internet access. If you’re having trouble recording, please contact us.

Voice and Message banking are often performed when preparing to lose the ability to speak, either due to a progressive disease or pre-operatively—often before any speech changes have occurred. In contrast, Project Euphonia is focused on improving speech recognition rather than recreating a person’s unique voice. Voice Banking, Message Banking, and Project Euphonia all require recording speech samples in as natural sounding speech as possible.

Just like our work on speech recognition, our work on voice recreation is very preliminary and not ready for external users yet. You can, however, try Acapela (or other providers) for voice banking, and/or store your own natural voice recordings (message banking) with tools like mymessagebanking.com.

We’d love to hear from you; press inquiries may be directed to press@google.com.

Adults 18 or older who speak English, French, Hindi, Japanese, or Spanish, and who have trouble being understood can sign up and subsequently record speech samples.

Yes! We have created an Android app designed to help those with non-standard speech communicate better. Find out more at g.co/ProjectRelate.

We are grateful for your interest! We have several SLPs on our team and partner closely with many Accessibility Allies, but realize that broader collaboration amongst this critical community is key to the success of our project. We’re currently scoping how we can best engage with you, optimally leveraging your experience and expertise while respecting your time. In the meantime, the best way you can help is to encourage the individuals you work with to learn about Euphonia and—if comfortable doing so—sign up to share speech samples to help accelerate our research.

Please fill out our SLP-focused survey if you haven’t done so already.

Participation in our research initiative is voluntary. However, we understand that these recordings take time and energy. The recordings you assist participants with will be reviewed by a speech and language specialist. If, after review by these specialists, your recordings are deemed suitable for Euphonia research, you will receive a gift card by email valued at $60 USD (or local currency equivalent) for 300 phrases recorded. An annual limit and other restrictions may apply. Examples of speech that are NOT suitable for Euphonia research include accented speech and standard speech.

We collect the minimum information required in order to:

• Ensure that participation in this research project is appropriate for the person (e.g., they speak English, they’re not a minor, etc.)

• Communicate with the participant (e.g., an email address)

• Comply with country-based laws about communications

We do not ask anything medically-sensitive (e.g., no condition-related questions). Those who choose to participate will only read and record non-personalized phrases.

Privacy is a fundamental aspect of our research protocols. Our signup form invites people to include their name, nickname, pseudonym, or an initial for us to use in communications. Upon request, we also disassociate your speech samples from the name you provide in the sign-up flow. You may also expunge records upon request.

Additionally, all of our SLPs and other team members work with identifiers (e.g., “Participant E12345”) when analyzing and discussing speech samples.

For more information, you may review the consent form that participants agree to when signing up.

Yes. As we want to ensure all participants are fully informed and willing, we respectfully request that:

• You discuss the project with them

• You mention the benefits (helping to accelerate research, receiving a gift card) while noting that—since this is early stage research—there will be no near-term improvement in automated speech recognition for them individually

• They explicitly consent to the sharing of their email address with Google for the purposes of engaging with Project Euphonia

Questions? Contact us