By now, everybody knows about Google expanding its robo-transcription function to include every video on YouTube. In fact, you may have read about it in a recent story in Speech Technology.
Well, that’s all fine and good. We here at SpeechTech Blog love transcription as much as the next guy. But every video? I mean there are some pretty insane/marginal/terrible videos on YouTube. Take for example Mannequin Dance Party. Do we really need this video transcribed? Do we really want it transcribed? Well, maybe. Who am I to judge? And, I suppose I shouldn’t be so negative. This next video, titled Talking Carrots definitely deserves transcription.
But my Speech Brother Eric. B and I still have doubts about Google’s ability to transcribe every video. So, that is why we are issuing the following Transcription Challenge to YouTube and Google.
My Speech-Brother Eric B. and I were enjoying a leisurely brunch here at The Home Office–reading the newspaper, sipping mimosas and eating our traditional Speech Tech Breakfast of poached eggs, kippers, rashers, fried bread and Jell-O–when we came across a story on the cover of The New York Times about speech technology.
Check out the above link, to read all the news that’s fit to print about Google Translate. We haven’t seen anything this high profile since The New Yorker Episode of 08.
According to an official press release, “GM Voices, the world’s largest producer of professionally-recorded voice for telephony applications, hit the recording studio (down the hall) and turned Asterisk open source inspiration into a rocking tribute to IVR, complete with jamming guitars, bangin’ drums, a white-hot harmonica solo and the angelic vocals of the “Call Center Girls,” customer service AND singing professionals.”
So, as we all count down to The Big Dance, check out Conference Chair Jim Larson dishing on some of his personal programme highlights from SpeechTEK Europe.
What are you looking forward to at SpeechTEK Europe 2010?
I believe that speech-enabled multimodal applications are the wave of the future, and so I’m particularly looking forward to Professor Wolfgang Wahlster’s Keynote presentation on research in European universities and company laboratories in the areas of speech, multimodal, and multilingual technologies so I can plan to use these new types of technologies when they become available.
In Developing Multimodal Applications, attendees will learn how to integrate web-based services to create “mashup” applications involving speech, and how to develop multimodal applications on a variety of mobile devices. I believe these approaches will be widely used to deploy speech-enabled multimodal applications that customers can use anywhere, anytime on nearly any device.
In our Panel Session: Advanced Speech Recognition Techniques and Experiences, developers will give first hand experiences about designing, implementing, and deploying call routing applications. I’m looking forward to learning from their successes as well as their mistakes in providing voice-based call routing.
The Using Voice Biometrics Session explains how to identify speakers and verify that speakers are who they claim to be. These promise to help minimise fraud and theft.
The European market place is multilingual, yet most of today’s speech applications only speak and understand a single language. SpeechTEK Europe has two Sessions on developing multilingual speech applications, how to conduct usability tests in multiple languages, and how to deal with cross-cultural attitudes about IVR and speech systems.
Other highlights for me at SpeechTEK Europe are sessions describing challenges in new IVR markets, key directions for unified communications and hosting/premise, and a panel of analysts predicting the future of speech in Europe.
It’s no secret that my Speech Brother Eric B. and I love Creepy Talking Robots. Just the other day, we were sitting around the old campfire, shelling peanuts and talking about all the Creepy Talking Robots we’ve profiled over the years: strange, vaguely sexual gynoids with feathered blond hair and trampy speech recognition engines.
And then it happened. We encountered a CTR the likes of which we thought did not–nay could not, could never–exist. Prepare, Speech-Heads, for the absolute horror and lurid nightmare that is Roxxxy.
This month, Daniel Ziv–vice president of customer interaction analytics at Verint Witness Actionable Solutions–tackles how understanding customer behavior is important regardless of the chosen communications medium in a pice titled Social Media Meets Speech Analytics.
Check It. As my Speech Brother Eric B. once said: “This one is gonna be a real barnburner.”
As all you Speech-Heads probably know, when it comes to product testing my Speech-Brother Eric B. was always the go-to-guy here at Speech Tech Blog. In his heyday, Eric B. would be testing five, six, seven speechy products every week, writing up reports, tuning, jotting down notes, talking to computers, etc.
Well, in his absence, I decided to try Ribbit Mobile. And let me just say this: Unlike a lot of speech solutions I have experienced, this one WORKS LIKE GANGBUSTERS!
I set it up on my cell phone (in a matter of minutes) logged onto the site on my computer and was good to go. My messages were transcribed almost perfectly and sent to me as a text message and as an email. And that was really just the tip of the iceberg. There were a lot of other cool features and functions and I heartily recommend the product.
Like I said, this is a Brief Summary. I have IVRs to navigate.
The questions about SpeechTEK Europe never stop around The Home Office. My Speech Brother Eric B. and I must field about 100 calls everyday:
What’s up with SpeechTEK University? What’s the deal with Conference Sessions? What do I need to know about Keynotes? What should I do with the precious moments of my one and only life? What’s poppin’ with Customer Case Studies?
How can your company provide the best security in an easy-to-use and cost-effective way? Find out by joining us for this interactive roundtable discussion about on-demand biometric speaker verification.
In today’s business environment customers, employees, and business partners are touching companies at multiple access points using devices that are rapidly evolving into multimodal communication platforms. At the same time, criminals have become alarmingly adept at compromising those devices and interactions. Companies are struggling to meet the competing demands of security and accessibility at a time when the global economy is pushing companies to their limits.
Join J. Markowitz, Consultants, Convergys, and Tellme for this roundtable discussion about on-demand biometric speaker verification. This roundtable will address how to:
Improve security, reduce operational expenses, and improve the customer experience associated with identity management and customer authentication.
Limit capital investment and leverage transactional pricing and fast implementation cycles to accelerate ROI.
Understand the design and technical challenges that companies should take into consideration when evaluating voice-based authentication strategies.
Evaluate the options available with pro/con trade-offs to help companies identify the right fit for their needs
I don’t know about you, but my Speech Brother Eric B. and I are always looking for MORE SPEECH! We’ll take it anyway we can get it: print items, blogs, web news, webinars and now PODCASTS!