Archive for the ‘speech recognition’ Tag
I stumbled upon this today during a podcasting workshop at Skowhegan Area High School with Rob Munzing. Rob’s a fan of David Pogue’s and since I’m a fan of Rob’s (he calls me “kid”), I decided to check out Pogue’s line of podcasts. You can search iTunes or go to http://www.davidpogue.com/
Pogue writes a technology column for the New York Times. At iTunes, my eyes immediately jumped to his January 20 video episode – Speech Recognition Comes to the Mac. I like his style of explaining new technologies. Check this out – it’s a clever overview of why he chose to adopt speech recognition software and why he’s particularly excited about MacSpeech Dictate.
Speech recognition software allows users to control their computers – including creating content – by “talking.” I put “talking” in quotations because we are referring to computers here, right? As of this point in the 21st Century, there are few machines that will respond to a human’s random ramblings. So, by “talking,” I mean the use of a series of intentional commands and clearly spoken words (even “continuous speech” recognition requires a level of consistency for highest accuracy). A simplified example of speech recognition is automated telephone systems…you know, the deceivingly affable voice on the other end of the phone…”…please tell me your order confirmation number…” “I’m sorry, that did not compute. Please press 1, followed by the # sign…” All joking aside, such systems are actually quite efficient. Otherwise, companies wouldn’t be using them to replace human operators, who they have to pay. If only we could reach a human being when we need/want to…
Speech recognition has been around for decades. Research began in 1936, funded and performed by the U.S. government in support of military and defense initiatives. It reached the commercial market in the early 1980s. But as with all emerging technologies, the most significant barrier to becoming mainstream has not been speech recognition itself, but the capability of the computer on which it runs. We’re talking computer speed and power. As that has increased, so has the accuracy and transcription speeds of speech recognition software. If the history of speech recognition piques your interest, check out a timeline of this technology.
The most recent obstacle, specifically for Apple users, has been the lack of a high quality and reliable product that is Mac-compatible. First there was IBM ViaVoice for Mac…then MacSpeech’s iListen. But many Mac users have longed for a speech recognition product that could compete with Dragon Naturally Speaking, a Windows only-compatible program. Well, last month a product that purports to do so was unveiled: MacSpeech Dictate.
MacSpeech Dictate, which requires an Intel-based Mac (such as the MLTI high school MacBooks), uses the Dragon speech recognition engine but is designed specifically for the Mac. Now, I don’t have any firsthand experience with Dictate (the initial shipments began February 15), but if it runs like Dragon, teachers and students should be quite satisfied. It is soooo critical, however, to not view speech recognition as a “jet pack” – it’s not for every student, and even if it is a good match for a kid, it may not be the sole solution for meeting their needs. Here are a couple of places to visit for objective news and views when it comes to the use of speech recognition by students:
Speaking to Write/Word for Word (reprinted article)
And here’s a nifty little PDF handout with tips: