Speech recognition news

It’s just over two week since we finished at Mobile World Congress and I’ve had a bit of time to reflect on the conversations I had with people that dropped by for a chat at our two exhibition stands – one located in the main section of Hall 2, the other in App Planet. There was certainly a healthy level of interest in how virtual assistants can be used on mobile devices – maybe this shouldn’t be such a surprise following Apple’s launch of Siri but it was really encouraging that so many people are recognizing the opportunity of using NLI as the basis of a new and effective speech enabled interface for mobile devices.

Every now and then someone writes an article on whether Back to the Future II got its predictions right. Well, judging by the presence of voice activated TVs and remote controls at the Consumer Electronics Show (CES) this year, I think we can add a tick in the box to the TV that you talk to. But as a report by Leo Kelion on the BBC website has pointed out, it is not without its flaws.

Artificial Solutions, the natural language interaction specialist that enables users to have a meaningful, humanlike interaction with technology, today announced Teneo Mobile - a platform independent alternative to Apple’s Siri technology and designed to enable organisations to create their own virtual assistant as a mobile app, regardless of platform, mobile device and even language.
Teneo Mobile allows users to interact with their smartphone or tablet using everyday, natural language - either spoken or typed - and provides intelligent, humanlike responses to queries in a user friendly, conversational style.

Jeopardy! competition is a popular US quiz show featuring trivia in various topics (history, literature, the arts, pop culture, science, sports, geography, wordplay, etc.), broadcasted since 1964. The show has a unique answer-and-question format in which contestants are presented with clues in the form of answers, and they must phrase their responses in form of a question.

From 14-16 February 2011 the first ever man vs. machine Jeopardy competition will be held. Human competitors will fight against Watson, a supercomputer that extracts knowledge from enormous amounts of data at a much faster rate than human beings or any other computer system is capable of.

Just wondering why I have never written about Milo, a virtual agent, conversational avatar, chatbot responding to voice, face and gestures. Very cool presentation.

Cassandra is our “Figure of Speech”, a title that nicely represents who and what she is.  ejTalk’s mission from the very beginning has been to focus on the basic elements of conversation instead of developing products to accomplish a singular task.  Most conversational agents today serve a particular purpose.  They talk about specific topics but do not communicate in very conversational ways.

ejTalk takes the opposite approach with Cassandra by laying a foundation for the basic principles of day-to-day speech that will one day allow her to talk about anything.  Take the example of a greeting: There are many ways to introduce oneself to another person.  We may say “hello” to a person we have just met, but to a friend we will say “hey” or “hi.”  It may seem like a trivial distinction, but to the human ear, the effect is highly noticeable.

Cassandra’s entry in the SpeechTEK Europe Avatar Challenge.

Do you ever ask yourself if you are going to need an umbrella today? Or: ‘Where can I get a green tea latte’? Or: ‘What’s up this weekend’? With a new free iPhone 3GS app Siri, you simply ask these questions by talking to your phone. Your question will be interpreted and displayed and suggestions will follow immediately. Very, very cool.

And when you feel like travelling this weekend, you can just continue your questions after ‘what’s up this weekend?’ by asking: ‘How about San Francisco?’, as if you would speak to a real person. Sire will recognize that you’re still interested in something to do for the weekend as it’s context sensitive.

Awesome! Also check out review below the fold showing more examples of how you can use

Steve Di Paola of the Simon Fraser University demonstrated a virtual agent system to Chatbots.org.  This system interprets real time emotions in input such as voice, and shows facial expressions and gestures.

During the 9th International Conference on Intelligent Virtual Agents (related to IVA Gala), Steve showed us how his parameterized facial animation system operates. His goal is to develop a system that can handle any face type, behavior, and voice to be used in games, movies and virtual agent worlds. Their current focus is on how emotions should be expressed.

In the first few minutes of the video, he demonstrated various types of faces. Steve grabbed his microphone and started talking. The avatar spoke similar to Steve and his (or her) lips were synchronized with his words. As soon as Steve raised or changed his voice, the avatar acted correspondingly, which was absolutely amazing. It widened its mouth when talking louder; it moved its eyebrow to emphasize what it was saying, and it responded to a drumming sound. Wow!

More explanation and images after break.

A revolutionary interactive 3DTV system is being created by De Montfort University Leicester (DMU), England, researchers. The €4.2 million (approx £3.7 million) project aims to develop a television that can recognise where somebody is sitting in a room and what they wish to view and interact with on their television.

Researchers believe it is a step towards truly interactive 3D video games where gamers use their bodies to control the action without the need for a controller. It could be the next step for Microsoft’s Project Natal.

The project, called HELIUM3D (high efficiency, laser-based, multi-user, multi-modal 3D display) is also exploring ways of allowing viewers who are watching the same television to each view a different channel at the same time and could even let them choose different viewing positions within the image.

For example, groups of people watching a football match in the same room could each pick the part of the stadium from which they would like to experience the action.

Microsoft’s Project Natal delivers a new addon to your XBox that allows you to play without controller. The system has a 3D camera that maps the exacts position of your hands, your fingers, your feet, your header, your nose, everything in a 3D map. This allows you to control the game with only your body, in great detail, and no controller needed. Furthermore, it recognises voice and faces and supports complex video chat.