Nuance Accuses TellMe of Patent Infringement

Nuance and Phonetic Systems Ltd say that Tellme is infringing on their patents. The claims of the patents cover features of directory assistance and call center apps that include “whisper” technologies and advanced database query techniques.

Read the article.

Speech Recognition-based Podcast Search

PodZinger has recently launched its podcast search engine. The search index is actually created by running podcast audio content through a speech recognizer. What is really powerful is how they display the search results. In addition to getting links to the podcast's rss feed, you get a list of time marks within the podcast marking where the term was used!

So for example, when I search with the term "speech recognition" and sort by relevancy, an NPR program from October 2005 discussing speech recognition technology shows up, with 50+ additional podcasts that recently used the term speech recognition. If I click on one of the time marks, the player (applet embedded in the html page) begins playing within the segment where speech recognition is discussed.

If you listen to podcasts and are interested in finding out who's saying what about specific topics, this is a very handy tool.

Read the article.

The Voice-Activated Grocery List

Here's an interesting announcement from SmartShopper, Inc. on a dedicated device that serves as your voice-activated grocery list:

http://biz.yahoo.com/bw/060209/20060209005019.html?.v=1

The device utilizes Nuance's VoCon 3200 embedded speech recognizer and allows you to utter items you need to pick up on your next grocery run. The device is also equipped with a thermal printer so you can print yourself a hard copy.

This is an interesting gadget, but why would anybody want to lug around an extra gadget when all the functionality you need to implement the same thing is available on a mid to high tier conventional mobile phone? For example, my mobile phone (~ $299 retail) is equipped with programmable speech technology that can easily and accurately recognize 2000+ grocery items. It also is equipped with a programmable bluetooth radio (JSR 82) which would make it pretty easy to connect to my $200 bluetooth-equipped HP Photosmart printer.

My mobile phone is also a network device, so if I wanted to quickly browse a few product reviews on that expensive bottle of wine I'm about to purchase while standing in the beverage isle, that can be accomplished without difficulty.

In addition to using bluetooth to interface with my printer it could be used to register which area of the store I'm in and automatically display items on my list that are in that area. For example, when I enter the baked goods/deli section the fresh bread and honey-baked sliced ham entries are displayed.

X+V Applications Galore

Last month we playfully took IBM's Igor Jablokov to task on the lack of X+V applications on the web that utilize the X+V capabilities of the Opera 8.0. Igor promptly supplied URLs to three X+V apps that you can try out for yourself. I quote from his comments:

Looking for a romantic restaurant to take your significant other? Ask for it at:

http://pvc002.austin.ibm.com/findit3/

Travelling on a quick day trip to Seattle and needing to know if you should bring an umbrella Try:

http://pvc002.austin.ibm.com/weatherit

Leaving the office in Atlanta and wanting to know which route to take? Better check with:

http://pvc002.austin.ibm.com/traffic

Since then, VoiceXML aficionado Jim Larson (of W3C fame) has posted over a dozen example X+V applications for X+V, authored by Jim's pupils at at Portland State University. While these are not polished commercial apps, let's not forget that if anybody can take an emerging technology and do something cool with it, its college kids. Be sure to get Opera 8 installed and try these applications out. Don't forget to provide Jim and Igor your feedback!