The new Voiceitt web app that can recognize free speech, is now available in English! Here is how to get access and use it.
Getting access to the new web app
To access the new web app, you will first need to provide us with your voice recordings, so we can customize the system to recognize your unique voice and speech patterns.
Once you have finished training, we will create your custom voice recognition model based on your training. After your model is complete, you will get full access to the web-app recognition feature, and you can start using it.
1.Log -in or Create an account
Log in to your account here. Click "log-in."
Create an account by Clicking on "create an account."
Provide the requested information and create your account.
Please be sure that only one user is associated with each email account. Our custom Voice recognition model is built upon an individual's speech patterns, therefore, individual users should have their own Voiceitt account.
Under 13 years old? We will send you a parental consent form for a parent to sign and send back to us. Once we get a signed consent form, we can provide access to the web app.
2. Training: Record Your Voice
Speech training is be done directly in the web app. Once you enter the web app for the first time, you will be taken to record several phrases which will be used to create your unique voice model.
Click on "General phrases."
now you will see a sentence written on the screen in blue.
1. Click on the "tap to speak" orange button.
2. Read the written phrase aloud
3. click again to stop recording, or wait for it to stop automatically
4. Click next
5. repeat these steps until you reach at least 1000 phrases.
Is a phrase too hard to say? You can click the arrow icon on the right to skip to the next phrase
Do you want to listen to your recording? You can click the speaker icon on the left to listen to your recording
Was the recording unsuccessful? You can click the "redo" icon on the right to re-record a phrase.
If you are using a PC, you can use the keyboard to train your Voiceitt web app:
start \ stop: [ space bar ]
Next phrase: N
skip phrase: S
listen to your recording: L
How many phrases do we need?
If you are new to the training process, please record ten phrases. We will listen to your sample recordings and then inform you via email if you are eligible. Please allow a few days for our speech specialists to review your recordings. You can then proceed to complete the training.
Once invited, we require a minimum of 1000 recordings to build your custom speech recognition model. You may want to do more recordings to get the Voiceitt web app to better recognize your speech.
Already done training in the Ensemble website?
If you reached 1000 recordings on ensemble, you might not need to do the training in the web app, and you might be eligible to get access right away! contact us at firstname.lastname@example.org to find out
If you haven't reached 1000 recordings on ensemble, but started recording, you may finish the 1000 recordings on the ensemble website.
Do you still have questions about training and using the web app? Please contact support at email@example.com
Voiceitt's training tips:
To speed - up the process of getting full access, you can follow these tips:
1. Make sure your environment is quiet and there is no noise or other speech in the background; only the account holder's voice should be present on the recording.
2. You can use a microphone for higher-quality recordings.
3. Read the sentence exactly as it is written, without adding or omitting words or syllables.
4. Feeling tired? that's okay! You can leave and enter training anytime, and all your recordings will be saved, and you can continue from where you left off.
If you are on the web app's homepage and would like to do some more training recordings:
a. Tap on the three lines on the top right
b. Click on "training phrases."
C. click on "General Phrases."
Training your personal phrases
You may also add customized phrases, and train them. This will help you expand the app's vocabulary, and help make recognition more accurate.
For example, you can add words, or names, the webb app is not recognizing (places, addresses, names of friends and family), or phrases you know you will be using, and want to ensure the web app knows them:
a. go to the training page (if you are not there yet): Tap on the three lines on the top right and Click on "training phrases."
b. click "Add personalized phrase"
c. Type in your phrases.
after every phrase you type in, click on the "+" icon to add it to your training list.
d. When you are done adding your phrases, click "Record"
e. Train your phrases the same way to train the general phrases. You can read about it here
You can add more personalized phrases anytime, just don't forget to tap "Record" and train them.
3. We will create your voice model and provide you with access
- We will review your recordings and start creating your unique voice model.
- Please give us some time to create your model. If any issues come up, we will contact you.
- You will be notified by email once your model is ready and your unique speech can be recognized!
How to use the new web app
Once your custom voice model is ready, you can start using your new Voiceitt web app.
Using the app in "conversation" mode:
Using the app in "dictation" mode:
You can toggle between the two mods:
on a PC, toggle the rectangular blue button on the bottom left-hand corner, or tap the "tab" button on your keyboard.
On a mobile device, go to the Menu (three horizontal lines) and toggle to dictation or conversation mode.
You can Also dictate punctuations marks!
If you say the name of a punctuation sign such as comma, period, question mark, Exclamation point, etc... Voiceitt will put it the sign you named.
For example, If you say: "Hello Exclamation point my name is Yubi period"
Voiceitt will dictate: "Hello! My name is Yubi. "
Listening, copying text, and sharing
You can listen to your Voiceitt web app dictation.
There are two listening options:
The text will be automatically be read aloud after it appears on the screen.
You can turn this feature on and off by clicking "read aloud."
In conversation mode: read aloud is on by default
In dictation mode: read aloud is off by default
You can change this by clicking on "Read aloud" on the side menu or tap the "enter" key on your PC keyboard.
By clicking the "play all" button, the whole text will be read aloud. Tapping the "P" key on your PC keyboard will also "play all".
Click on "copy text" in the side menu or tap "C" on your PC keyboard to copy the text that was dictated. You may then paste it to wherever you like.
Deleting the text
Click "remove text" on the side menu or tap "R" on your PC keyboard to delete the text on the screen. On a mobile device, click the trash icon on the top left.
click "Share" on the side menu or tap "S" on your PC keyboard to share your text to social media, email, and more.
Keyboard shortcuts are available while using your Voiceitt web app on a PC.
start \ stop recording: [ space bar ]
Toggle between modes : tab bar ->|
on \ off read aloud option: Enter ↩️
Play all: P
Remove text: R
Copy text: C
Tip: The app can only listen for about 8 seconds at a time. Try to divide your message into short phrases, and wait for the app to write down what you say before speaking again.
Who is the ideal Voiceitt web app user?
The Voiceitt web app supports users with non-standard speech. This means Voiceitt can be used by individuals with various diagnoses that may impact their speech. Some of those diagnoses are Cerebral Palsy (CP), Downs Syndrome, autism spectrum disorders (ASD), Parkinson's disease (PD), amyotrophic lateral sclerosis (ALS), multiple sclerosis (MS), muscular dystrophy (MD), Huntington's disease (HD), deaf or hard of hearing and more.
In a nutshell, the web app is best suited for:
- Individuals who find that standard speech-recognition software does not always understand them due to their speech impairment.
- Individuals with impaired speech who are generally understood by familiar listeners but less understood by unfamiliar listeners.
- Individuals with intact language who say what they intend to and speak in full sentences.
- Individuals who are interested in using speech recognition technology for communication.
- English speakers
Before accessing the web app, we will have to listen to a few recordings to be sure you are a good fit. You can read more about this here.
What does the Voiceitt web app do?
The Voiceitt web app is a speech-to-text augmentative and alternative communication (AAC) software that allows people with speech impairments to use their voice for communication. This system was built to understand non-standard speakers, such as those with motor-speech disabilities, and "translate" their speech to text and automated voice output. It listens to what you say, then writes it down on the screen and can play it aloud. The dictated text can be copied and pasted somewhere else, so you can dictate your emails, social media posts, private messages, etc.
What can the new web app understand?
The new web app is built to understand free and spontaneous non-standard speech even if what you are saying was not pre-trained.
Note: the recognition might be partially accurate at first, but the more you use and train it, the more accurate it becomes. We constantly work to improve our technology and welcome your comments and feedback. Please reach us at firstname.lastname@example.org.
What devices will the new web app work on?
The web app opens on any device connected to the internet and can open a web browser.
Do I need internet access?
Yes. The Voiceitt web app works on the web, so internet access is required.
Do I need to train the Voiceitt web app?
Yes, a minimum of 1000 phrases must be trained because every person has a unique speech pattern and voice. The system needs to be customized for every new user. The training is used to create a voice model for every user. You can read more about training here
How long does it take to create my voice model?
Please allow two weeks for Voiceitt to build your custom recognition model. You can always follow up on the progress by contacting us at email@example.com.
How much does it cost?
The web app is free to beta testers who have recorded a minimum of 1000 phrases. Beta testers play an integral role in improving voice recognition equity for all. By training phrases, our voice recognition technology improves while also reducing the amount of phrase recordings required to build a viable model.
Can I listen to or copy the dictated text?
Yes. There is an option for every sentence to play aloud automatically after you say it. Once you have finished dictating your thoughts, there is an option to play the entire text aloud by clicking "play all." You may also copy the text and paste it to an email or word processing software.
What is the difference between the web app and the Voiceitt iOS app?
* While the Voiceitt iOS app can only understand pre-defined and trained phrases, the new web app is designed to understand spontaneous conversation and to understand phrases that were not pre-trained.
* While the Voiceitt iOS app can connect to Amazon Alexa and allows you to control your smart home, the web app is for communicating your thoughts. It does not connect to Alexa or any other smart assistant however, it can still be used to control smart home assistants through the traditional route of web app speaker output to your smart assistant microphone input.
The Voiceitt iOS app and web app are two different products. An individual can use both technologies without interference from each other.
How many users can be registered under one email account?
One! Since the Voiceitt web app is custom-built for each individual, everyone who wishes to experience this technology must create and use their own email account.