I’m always looking for new functions and ways to use different features in visual assistance applications, and first downloaded the Envision app a few years ago when my friend was raving about its text recognition features at a low vision accessibility event we were both attending. When Envision became free for all users in 2022, I began recommending Envision and Envision AI more often for users looking for a visual assistance app that would work on multiple platforms and support helpful features like file upload, multiple languages, and recognizing faces. Here are my tips for using the Envision app with low vision for a variety of tasks.
The Envision app (also known as Envision AI) is a free visual assistance app that uses machine learning and artificial intelligence to recognize text and other types of visual information, including colors, products, people, and more. Envision is designed for blind and low vision users and offers multiple accessibility features, including self-voicing descriptions (no screen reader needed), large font sizes, dark mode, and other accessibility features that make it easier for people to get the information they need.
Unlike other visual assistance apps that are exclusive to one platform or operating system, Envision is available for both Android and iOS devices. On the Apple App Store, Envision is listed under the name Envision AI, but the features between the Android and iOS app are identical.
Before using the Envision app for the first time, users are prompted to log in or create a free Envision account in order to use the app. New users will need to connect their Google or Apple account, or provide an email address and password to create an account.
The default view for the Envision app is a live view of the device’s back camera with a series of buttons on the bottom of the screen that are labeled with various functions/features.
To activate a function, users position their device camera over an object, text, or other area of interest, and select the button of their choice to either read or listen to a visual description/transcribed text. Once the user is finished, they can select the button again or select the “back” button (if available) to return to the default view or choose another option.
There are a few different ways to input or upload content for Envision and Envision AI, including:
The Envision App itself is available in over 30 different languages and local translations. However, the Envision app can recognize text in more than 90 different languages, including several indigenous languages, endangered languages, and even dead languages such as Latin and Ancient Greek. A full list of supported languages are linked below.
For text-based features, users can change the reading language for increased accuracy by selecting the Reading Language option at the top of the screen and choosing the language they want to use- this does not change the language of the Envision app itself.
The Envision app and Envision AI app offer the following functions. Depending on customization settings, some features may be hidden under the More Features menu, though users can customize the display order of functions and rearrange items as needed.
Instant Text reads text that is visible within the view of the device camera. This can include environmental text like signs, flyers, room numbers, or shorter documents that don’t need to be saved for later. Both handwriting and typed text are supported.
Scan Text reads text that is visible in a photo taken by the user within the Envision app. This can include longer single page or multi-page documents like a class reading assignment, letter, or handwritten messages. Before transcribing text, users are prompted to take a picture of the document they want to read, and the Envision app then uses character recognition technology to identify text. Once the text has been identified, users can then read the text in a simplified display, listen to text read out loud, or export text by selecting the Share icon at the bottom of the screen.
Users can adjust the text layout detection in the Settings app.
Library stores files that have been uploaded to the Envision app and are saved locally on the device, not uploaded online. Users can upload files from their device and read them in the Envision app, with the option to save the file to the Library for later access. If the Envision app is deleted, users will lose access to files uploaded to the Library.
Instead of using the device camera, users can import files from their device or import an image to receive a description or read text. Items can be saved to the Envision library for later use, and can also be translated using the translation option at the top of the screen.
Scan Barcode identifies product names and other details, and provides the option to view additional product options on Google.
Describe Scene takes a photo of what’s in view of the device camera and provides a simple visual description of various objects in the image, as well as people if relevant. Images can also be saved to the device photo gallery.
Explore uses the live device camera to identify supported objects and people that are within the view of the device camera.
Find Objects can locate common items that are within view of the device camera, prompting users to select the object they are looking for. Some examples of objects include:
If Envision detects an object, the user will receive an audio cue that the object is within view of the device camera.
Find People determines if there are people nearby, or if a specific person is within view. If people are detected, the user will receive an audio cue, and the person’s name will be spoken if they have previously been added to the Envision app.
Teach Envision provides users with the option to recognize a face by taking five photos of a person’s face and adding their name to the list of recognizable people.
It’s worth noting that this information is stored locally on the device, uploading a user’s face does not make their name and likeness available to anyone who uses the application. So if I trained Envision to recognize my professor, my friend’s Envision app would not automatically recognize my professor unless my friend also used the Teach Envision function for them.
Detect Colors tells a user what color an object is by using a sensor in the center of the screen. For items that have multiple colors, only the color that is in the center of the device screen will be identified- users will need to move the device around to get additional color information.
Within the Settings menu, users can select between Standard and Descriptive color options. Descriptive offers additional color shade information and color names.
Scan Accessible QR is designed for scanning QR codes that are specifically designed for accessibility- accessible QR codes have a border around them and can be detected from far away. Some products have accessible QR codes on their packaging, and Envision can identify the names of these products or other information as needed.
Here are some examples of ways I have used Envision on my Android phone and Envision AI on my iPad in various environments and contexts
Envision isn’t the only visual assistance application I use, and there are a few tasks where Envision isn’t particularly helpful and I prefer to use something else. This can include: