Working the Spoken Word

If you're a student with cool ideas about accessible user interfaces - using speech recognition or synthesis, for example - consider the special award at the Imagine Cup . At this year's competition, Microsoft is calling on young programmers, artists...

Yeah, yeah, the year in review, what a crushingly unoriginal idea for a post. But wait - this is worth it. 2007 was a huge year for speech recognition products at Microsoft. I think we'll look back on it as a real turning point. Here's how it shaped up...

According to a report in the Seattle Times yesterday, 21 out of 30 serious users of the Roomba vacuum-cleaning robot give their machine a name. More than half assign it a gender (male) and others have been known to dress it up. What kind of human-machine...

The ability to extract the audio for an entire call (both prompts and recognitions) from the Speech Server 2007 logs is a really useful feature for a number of analysis and tuning scenarios. Since the topic has surfaced a few times on the Beta newsgroups...

Sullen student, bored by x-rays of terrible chest diseases, is mesmerised by speech recognition , mutters that's so cool . Can't argue with that. (And can't resist real stories with a whiff of The Onion .)

Speech recognition in Vista works well for me. Dictation accuracy is very high - especially since I flicked the switch to train it on my emails and documents - and the correction experience is smooth and efficient. But I hardly use it. I find it very...

We've all had frustrating experiences with speech recognition systems, and as a race we're not beyond punishing virtual beings the same way we would punish people. So, what to do when that voicebot won't behave? Teach it a lesson! Here are some tips on...

I've just switched groups at Microsoft, and for the first few days it felt strangely quiet in my new office. The acoustic background hadn't changed much: I could hear voices in the offices nearby, people still passed by in the hallway chatting (some stopped...

Respect your elders - here's a video of the state of the art of speech recognition research at Stanford in 1968. I thought it was a hoax at first, with the synth music and the board titles, the fuzzy waveforms, and somebody actually acting out "I scream...

Sad news from Cambridge this week- Karen Spärck Jones has died, aged 71. I took her classes on NL systems in my M.Phil,, and she was the advisor on my thesis (a text generation program to describe images). Karen was active in developing real systems since...

(Tap-tap. Is this working? OK. Hh-hmm.) We have shipped the Speech Server 2007 Public Beta! It's available here: http://www.microsoft.com/downloads/details.aspx?FamilyId=4F4D3AA4-8223-406C-B74F-DB2DE928D8B2 Since Speech Server is now technically...

Continuing an irreverent dictionary of voicebot vocabulary: yourcallisimportanttous , cl. Obsequious attempt to make the caller feel better about a dud place in the cattle queue. Actually means the opposite. Often followed by pleasestayontheline...

Big time ! This kind of brainpower and expertise, on this kind of scale... the future is here. Welcome, Tellme!

Reading Nicholas Carr's dissection of the blogosphere this morning as "a vast, earth-engirdling digestive track, breaking down the news of the day into ever finer particles of meaning (and ever more concentrated toxins)" I am inspired to do my bit as...

"He never works when he's in a bad mood..." says the the Boston Globe in a profile of successful voice talent Tom Glynn. And given the amount of thought and work that he puts into every single prompt, it's hard to fault that.

IBM engineer and blogger Michael Tolva contacted Richard Powers in disbelief about the dictation of his recent work (see my last post ). Here's Mr Powers' reply about his correction UI: ...the combination of stylus and speech is far more fluid and...

Speaking in last Sunday's New York Times , novelist Richard Powers tells how he has used speech recognition on his TabletPC to compose all his writings of recent years. It's a fine article - you could quote him almost anywhere on the obstacles of the...

As planning for the next wave of speech server technologies gets underway, I'll try to post more on the broader context of Unified Communications. I'd love to get your feedback on the communication scenarios, applications or features that you'd like to...

Not sure why I'm fascinated by this study . Maybe it's the 98% slot-filling rate. Maybe it's the notion that out-of-grammar responses are 7 times more likely for types of pain than for body parts (yeah, think about that...) Maybe because it's a well-written...

Ouch. The Consumerist blog picks up a graph in a Fast Company article that puts speech recognition technology at 4 on a descending scale of 10 (humane) to 0 (inhumane) , then the commenters really get stuck in. The comments make fascinating reading...

Here's an inside look at the 'voice talent' industry - the people who provide prompts (and thereby branding) for companies with automated systems - with the following data points: Perhaps half of all Fortune 1,000 companies still rely on their own...

"Fancy-pants" speech technology gets the blame for... a database failure? Paul McNamara, an editor at Network World, recently blogged about a couple of calls he made to his healthcare provider's customer service, which is fronted by a speech recognition...

One in an ongoing series on the all-too-common vocabulary found in Voicebot systems today. pressorsay n , cl. A request for the caller to choose an option by entering a number on the keypad or by speech. Effective in dehumanizing and demeaning the...

Ah, the old pizza application. I'd like a large margherita, please . The classic sample app for speech recognition dialogs. Like 'Hello World' but extended into the to and fro of conversation. Great for illustrating prompts, grammars and dialog nuances...

In the same week, two outspoken champions of the consumer present opposing views: - Paul says Connect to a human whenever possible (via Lorna). - Peter says Talk to a person? Ugh. What do you think? "If a human operator is available when a...