Steve Kelley

Can you imagine taking on the big players in the world of Artificial Intelligence apps, like Microsoft's Seeing AI or Google's Lookout for document reading or product identification chores? AudiVision is doing just that with their new app, currently available from the Google Play Store. Like Seeing AI and Lookout, it offers a free version. Unlike them, it is funded with banner ads which appear at the bottom of the app display. These are relatively easy to ignore and users can remove the ads by purchasing the app for an annual subscription of $10.

For now, AudiVision is only available for Android, and users will find it in the Google Play Store. According to the developers, though, an iOS version is on the way, and should be available in January of 2023. Like Google's Lookout app, AudiVision has several modes which use the Android phone or tablet's camera to read text and identify objects. There are some similarities, like Text and Document modes, and both use artificial intelligence (AI) to identify various objects in the environment. AudiVision has a number of unique features which make it stand out.

For example, one really convenient feature in AudiVision is the addition of speech commands to open the various modes, called Tabs. Neither Lookout or Seeing AI uses speech commands. Each mode provides a microphone icon in the lower right corner. Tap the microphone and speak the mode to be opened. Speech commands can also be used by touching anywhere on the screen for a second. When the screen vibrates, just say the command. The commands are listed in the Help mode and can be read even without TalkBack turned on. When the Help Tab is opened, the user is told to "single tap to read instructions." A single tap also pauses instructions. This is a nice touch for the low vision user who may not be using TalkBack all the time.

AudiVision has a total of 10 tabs which appear across the top of the screen. The first 7 of these tabs are various modes for recognizing text or objects in the environment.

AudiVision Modes or Tabs

Here's a breakdown of the 10 tabs and a brief description.

  1. Smart Text- reads any text the camera encounters. This could be useful for reading mail, something on the computer screen, or scanning signs on a walk to see what's nearby.
  2. Document- is better for reading a page in a magazine, book, or paper document. Hold the camera above the document and follow the prompts. A picture is automatically taken when it is completely in view, or the user may take a picture by touching the screen. AudiVision processes the image quickly and the text on the screen can be magnified as needed. Maximum magnification on the reviewer''s Galaxy phone was 3 character per line. Text can be played and paused using the controller at the top of the screen if TalkBack is not turned on.
  3. Find- uses an edit field or speech to select a particular word to be found. With the text the user wants to find entered or spoken, moving the phone systematically around an area will conduct a search for that word. This might be useful to locate a nearby office, a product in the grocery store, a business on the street, or a nearby document on the desk. When the word is found the app vibrates and sounds an alarm.
  4. Magnify- is like a handheld video magnifier. The magnifier tab is a feature not available in either Seeing AI or Lookout. This mode includes at least eight magnification levels, by pressing the plus (+) or minus (-) buttons. There is also a button to turn on the light, and a button to reverse the viewing mode to a light on dark negative mode which will provide more contrast for some users.
  5. Expiry Date- to help locate and read out the expiration date on a food item. When this tab opens, users hear, "This is an experimental feature. Please reconfirm results with human agent." The results during the review were mixed. The expiration date on one item in the refrigerator was located and read quickly. If a positive identification is made, it is read aloud and shared to a blank page. The text can be copied or shared with another app. Several others were not found even though the camera appeared to be over the expiration date. It is worth noting for those using the app visually, without Talkback on, that the results seemed better when using Talkback. Talkback seemed to be reading the various attempts at reading the expiration dates. Even though these weren't always confirmed, or sent to a blank page, hearing the attempts sometimes provided enough information to get the date or a portion of it.
  6. Light- another feature not found on either Lookout or Seeing AI. When this tab is selected a continuous pulsating sound plays and the pitch is higher as it approaches a light source, and deeper as it moves away to a darker area.
  7. Mask Detector- lets a user know if people nearby are wearing protective masks. Like the Expiry Date tab above, better results were achieved when Talkback was on. People were identified in the camera view and the app reported whether or not they were wearing a mask. This information popped up on the screen and quickly went away so it seemed easier to listen for Talkback to read these results. For the review, one person modeled a mask and the app detected right away whether they had a mask on or not.
  8. Settings- includes several toggles to mute or unmute voice feedback, turn auto flash off or on, and to turn on auto capture for the Documents tab.
  9. Buy- this is where a user can purchase the app with an annual fee of $10 or a one-time fee of $100. The description on the Buy tab's page did not provide the purchase price and indicated that by pressing the Purchase button, the app could be purchased. No Purchase button was visible, however on the Galaxy phone used in the review, unless it was hidden beneath the banner ad at the bottom of the screen. The purchase information came from contacting the developer from their website.
  10. Help- provides some basic instruction on using the voice commands and a brief introduction to the various tab functions.

How does it Compare?

Both the Smart Text tab and the Document tab seemed to perform as well as either Lookout or Seeing AI. Results were responsive and quite accurate for comparable targets, like addresses on mail and magazine articles. Expiry Date and Mask Detection are both innovative and useful features. Although experimental, the Expiry Date function seemed useful on several food items.

In addition to the Text and Document features on Lookout, comparable to Smart Text and Document on AudiVision, this reviewer finds the Food Label and Currency really handy. AudiVision, at the moment is not going to replace either Lookout or Seeing AI—it's going to be a great addition to either, at a very affordable price. For more information, check out the AudiVision website or take a look at the AudiVision (originally called Audible Vision) review on the Blind Life.

This article is made possible in part by generous funding from the James H. and Alice Teubert Charitable Trust, Huntington, West Virginia.

Comment on this article.

Previous Article

Next Article

Back to Table of Contents

Steven Kelley
Article Topic
Product Reviews and Guides