The way they refer to it is seriously misleading because it makes you think that voiceover somehow knows to automatically label unlabeled UI elements, and they stay labeled, while they're actually talking about screen recognition, which is available on SE2, and which is more like OCR.
When you enable it in voiceover preferences/voiceover recognition, and then use the gesture or rotor to turn it on/off for an app, it basically performs an OCR scan of the window, and lets you interact with the result just like you would a normal app, except now certain elements may be visible and clickable.
Comments
Yes, but that title is misleading
The way they refer to it is seriously misleading because it makes you think that voiceover somehow knows to automatically label unlabeled UI elements, and they stay labeled, while they're actually talking about screen recognition, which is available on SE2, and which is more like OCR.
When you enable it in voiceover preferences/voiceover recognition, and then use the gesture or rotor to turn it on/off for an app, it basically performs an OCR scan of the window, and lets you interact with the result just like you would a normal app, except now certain elements may be visible and clickable.