Re: labeling graphics?

Josh Kennedy

I posted two or three issues already over the past several years and nothing has been done about adding a graphics labeler or an object labeler or a frames manager to NVDA for making an ad on to do so. I think that NVDA should start to get away from scripting, and start using artificial intelligence and neural engine technology such as is used in the new voiceover for iOS 14. That way NVDA could start to read more things automatically and would rely less on scripting which can easily break when an application version changes The only other screen meter that I know of which used a primitive kind of artificial intelligence to read the screen in which not which did not rely on scripts was ASAP, automatic screen access program for Microsoft DOS made by micro talk. that screen reader was discontinued at around the year 2000 because windows took over. But for its time it was the best screen reader for DOS, or the disk operating system operating system which was available and which was ever made. I hope soon that somebody will at least make an ad on for NVDA which will use the free tensorflow neural Engine technology to allow NVDA to intelligently read the screen even if buttons and other controls may not be accessible in the standard way using standard accessibility measures. Voiceover for iOS can do this even if buttons and other objects are directly drawn onto the screen it can still identify them using artificial intelligence technology. It is amazing! I hope NVDA could do something similar in the future.

Join to automatically receive all group messages.