Monday, December 10, 2007

I wonder if you can hear my blog


Link to Podcast (RSS feed) for this blog

Friday, December 7, 2007

Email Bilingual Assistant

Mike Bilbrough has written, in HTML and JavaScript a very interesting program which assists in writing emails in the Business domain. It is still under development, and the author hasn't even given it a name, but I will keep updates on this issue.

The great thing about it is that it is a bilingual tool. It includes a lot of applications, so while writing your email in the foreign language (English/Spanish) you can insert words from the dictionary, check the spelling and the grammar rules. It is very complete and user-friendly. The new version will include a TTS application which will read your emails as you're writing them.

The TTS software is a link to FreeTTS 1.2, a Speech Synthesizer written in JavaTM Programming which perfectly integrates with Mike's email assistant. See more on FreeTTS 1.2 in http://freetts.sourceforge.net/docs/index.php

Wednesday, July 25, 2007

Interesting Voice Samples from Real Text-To-Speech Applications


I would like to share with you these really good voice samples that I have gathered from different speech applications.
AT&T Natural Voices
NeoSpeech
Acapela
RealSpeech

You can find more voices in Nextup.com

Speech Recognition in Windows Vista?

Windows has been long in the field of Speech Recognition. They, however, have been happy enough in the past without resorting to this sophisticated form of technology. They could make millions without really putting much innovation in their products. This video is representative enough of how technology can be kidnapped by the powerful and be administered little by little till it is super profitable. In short, does society really give big leaps into the future? It is not that Speech Recognition is working at full speed, but sure it can do more than this videopodcast ironically shows.

Tuesday, July 24, 2007

Joint Writing and Speech Recognition for Content Channels

In my search for Speech Applications in the field of Instructional Technology, I have come across an interesting study on Content Channels. Content channels refer to the various sources of information in some context. In this study, the authors, Richard Anderson, Crystal Hoyer, Craig Prince, Jonathan Su, Fred Videon, and Steve Wolfman explore the possibilities of Speech Recognition, Digital Ink and Slides in a Technology-Enhanced classroom. The idea is just marvelous, awesome: imagine a Distance Education situation, or a class for the hearing impaired, with the combination of a Tablet PC, Speech Recognition and a simple Slide Presentation Software, there is so much you can improve to keep interaction with your students! I especially appreciate the writing-speech recognizer. Just a way to keep the teacher's voice and writing hand to hand. The study can certainly provide some insight in this interesting field of research: the teacher is talking and writing on the Tablet PC, and these two interfaces are combined to provide an enriched input for the student. Amazing. Have a look at this slide presentation: click here

Thursday, July 19, 2007

Human vs. Machine Communication

Many people believe using speech technologies interfaces to communicate with machines is but just the most natural way of human-machine communication. The reality could be very different. Even though Speech Technologies do indeed await an increasing development, it is also necessary to understand when such an interface is necessary or when it is advisable to keep the keyboard and other more traditional computer interfaces. Technology and human development should evolve at the same pace. Many times, technology is posing unnecessary risks and demands on people that we are not ready or willing to accept.