Contribute to Common Voice


The Common Voice project is Mozilla’s initiative to help machines understand how real people talk. We want to create advanced voice technology, which anyone can access for free, in order to drive innovation. 

To create voice systems, large voice data sets are needed. The Common Voice project aims to create an open voice data set to make it possible for everyone to create and experiment with voice recognition technology.

How to contribute

This contribution activity can be on your own or with friends/your community. Why not try them all?

Make a difference by sharing your voice and reviewing data clips. You can easily do that on the common voice website.

A quality dataset benefits from a diversity of accents and recording qualities, which is why we need contributors from multiple locations!

How to Donate your voice  Visit the common voice website homepage and create an account. Click on „Speak.“

Click on the microphone and read the sentence out loud. Click the red stop icon when you’re finished. Keep recording until you’ve reached at least five sentences and then click submit.

 

How to review voice clips  Go to the common voice website homepage and click on “Listen.”

Listen to the voice (try not to read the screen). Now read the sentence and compare it with what you heard. If the voice clip matches the sentence, click “yes.” If the voice clip doesn’t match the sentence, click “no.” (Note: different accents are fine, and encouraged!) Repeat.

Contribute with your friends, family or organize a larger event for a giant Common Voice contribute-a-thon. Here is a POSTER you can use for your event!

Workflow of a smaller event (less than 10 people)

Introduce participants to the Common Voice project (10 minutes)

Introduce the participants to the common voice project. Talk about why it’s important and show them how easy contributing is with just a smartphone. Here are some SLIDES that might be helpful. You can also consider visiting the website voice.mozilla.org together.

 Have all your participants create an account on their device (10 minutes) Spend 15 minutes donating voice clips (15-30 minutes)

To start, click on:

P.S. If the space is small or loud you may want to split the group in two, with one half donating voice clips and the other half reviewing clips, switching after 15 minutes.

Review other people’s voice clips (20 minutes)

After you donated 15 minutes (225 clips) start reviewing other people’s voices using the “Listen” function on the Common Voice homepage.

For a large group, you can have the group split up into pairs. One person will play the recording on their device and the other person will listen (without reading the screen). The listener will repeat what they heard to confirm if the voice clip is accurate.

After 15 minutes, ask the pairs to switch roles.

Keep track of participant progress.

You can easily transform contributions in a game! Make sure your participants have an account set up so everyone can keep track of their contributions using the counter at the top of the website:

You can:

Announce when people pass 25, 50, 100 clips. You can even create a leaderboard in your space (advised for smaller groups) Have prizes for people that hit their goal first (advised for smaller groups) Have prizes for people who  record the most clips (advised for larger groups)

 

Hand out prizes!

Prizes can be something like stickers or even as simple as a shout-out at the event.

Wrap-up and share (5 minutes)

The activity owner will ask participants to share their progress and broadcast the event using #CommonVoice on social media.

You can set up a booth at a conference or in a large public space to collect as many voices as possible and to spread the word about the common voice project.

We recommend setting up the booth with at least one or two devices using Firefox in full-screen mode with the Common Voice website pre-loaded. You can use this Common Voice POSTER and Instructions DOCUMENT to help teach participants how to contribute.

Spread the word:

Use your contribution booth to explain to participants how their voices can help create a public dataset that benefits everyone.

Examples:

Hello! How are you doing? Did you know you can help make voice recognition technologies work better for everyone by donating your voice? Hi! Do you have a minute to talk about data privacy and voice technology? The Common Voice project is Mozilla’s initiative to help machines understand how real people talk. Hello there! Do you have a minute to help our open-source voice technology initiative? Collect voices:

After this initial approach, quickly introduce people to the Common Voice project and invite them to donate a few minutes of their time to contribute. Even three minutes of recording will help!

Stay in touch

Help us find more people who want to donate their voices! To share on social media – “I am helping teach machines how real people speak. I’ve just donated my voice at https://voice.mozilla.org #CommonVoice”

If you run into any problems – ask questions in the Common Voice Discourse or join the conversation on the Common Voice Matrix room.

Have feedback about this activity? File an issue on GitHub or leave a comment on Discourse.