Saturday, January 28, 2012

Screen readers ("Evi" vs. Siri)

Comparison of screen readers - Wikipedia, the free encyclopedia

Siri at Apple iPhone 4s has finally managed to make computer voice communication mainstream.

How about "reading text?"

While this should be much simpler task than recognizing speech,
there are no many usable tools available...

There is one "special" among those tools: JavaScript based "webanywhere"
by University of Washington.

It works better with simple text, without web links.
Here is one such example:
http://www.gutenberg.org/cache/epub/11/pg11.txt
(Project Gutenberg's Alice's Adventures in Wonderland, by Lewis Carroll)

There is also a plugin for Firefox browser "Text to voice".
While quite usable, the quality is not quite ready for mainsream...

Why did I start to investigate this?

There is a new competitor to Siri, called "Evi"
Story at MIT Technology Review
that I in fact listen as podcast, generated by "audiodizer".

Evi one-ups Siri for iOS and Android @ slashgear


The "Evi" is created by TrueKnowledge
and is available for free on Android, and for .99 on iPhone.
Why the difference in price?
Android has Text-to-speech API embedded, and for iPhone license from Nuance is needed...
TrueKnowlage is using Semantic (Web?) technology for finding information...
Siri is using AI

I thought something similar can be done on Windows without much trouble...
Windows also used to have "Agent" tools, until Windows 7 when it is apparently deprecated...
Now, Microsoft offers "TellMe" tools.

Nuance, provider of voice technology for iPhone 4S, also offers text-to-speech tools,
as well as demo web site: Vocalizer 5
The demo accepts max 140 characters (like Twitter), and the quality is quite good,
plus it is fun to select from a few dozen international languages, that read English with accent :)
But Vocalizer is apparently sold only in some special way, for solutions...

Maybe Android is the best platform for this feature after all..
It just need a bit more work on apps.
"Evi" just responded (by voice) on an Android phone:
"I'm really sorry. Due to overwhelming demand, my servers are overloaded..."
No much real challenge for Siri and Apple just yet... :)

htmlUI: HTML5 & CSS3 Demos

htmlUI « HTML5 & CSS3 Demos

a collection of homegrown HTML5 and CSS3 experiments