[qi:83] Most of us will agree that these days trying to do anything other than making plain vanilla phone calls and sending short messages on a regular 12-keypad mobile phone is an ordeal. The triple tapping gives the thumb muscles a complete workout. No wonder it needs a new interface.
Vlingo Corp., a Cambridge, Mass-based start-up thinks it has the answer – voice-based interface – and has introducing an eponymously named application to showcase its technology. The application allows mobile phone users to speak and look-up information such as local restaurant listings to film schedules.
The company is the brainchild of speech recognition technology veterans Mike Phillips and John Nguyen. Phillips started Speech Works that was acquired by Scan Soft, which acquired Nuance and then renamed itself, Nuance. It has raised $6.5 million in venture financing from Charles River Ventures.
Here is how it works:
Vlingo’s voice-powered interface lets users speak or type into any vlingo-enabled text box on their mobile phones. Once the application is installed and started, you can say, sushi restaurants in San Francisco, the app sends your voice clip over the mobile data channel to a datacenter, where the voice is put through speech recognition software running on a server.
Server converts voice to text, and the results are sent back to the phone. Voice-turned-into-text then populates the search text box, and you can hit search and find local listings. The application can work with other applications such as music stores preloaded on the phones.
I was fully expecting some seriously latency, and was surprised by how quickly the results came back over an EVDO enabled Sprint phone. The system also learns your voice patterns, and you can fix any errors by using the keypad, and system learns those as well.
Vlingo is a J2ME App and as a result it can work with other applications on the phone and populate their text boxes as well. We saw it working with the Sprint Music store, in a demo we go from
Dave Grannan, chief executive officer of the company.
While Vlingo, the application is a showcase, the company is hoping to sign-up mobile carriers and handset makers. He said that company’s technology is going to be particularly attractive for carriers.
Mobile phone companies have spent billions of dollars developing fast networks, and cobbling up services such as music stores, only to be stymied by the challenge of entering complex information through a limited a 12-keypad system, which we can use to enter the data. Vlingo is going to have some carrier-related announcements later this year.
Vlingo faces quite a few challenges, none bigger than Microsoft, which thanks to its $800 million-plus acquisition of Tell Me Networks, has made its intentions clear that it wants to dominate the voice-mobile interface business.
Most importantly, mobile phone users have to feel compelled to buy into this offering, which is going to result in some extra spending on data services.
Vlingo Corp Fact Sheet
| Website | Vlingo.com |
| Location | Cambridge, Massachussets |
| Cofounders | Mike Phillips, Nguyen |
| CEO | David Grannan |
| Financing | $6.5 million |
| Investors | Charles River Ventures. |
| Business/industry | Mobile, Telecom |
| Competitors | Microsoft, VoiceSignal. |
The problem is the lack of context coming from the device itself. So the solution is to provide a way that requires virtually no data entry on behalf of the customer. Just because I can talk to the phone doesn’t solve the problem – the web server still has to resolve the request and invariably it requires more context to do that on a mobile device. Any app can support voice – the trick is to support voice and other real time contexts such as location, device information and personal information in a way that integrates into the browser so it’s available for the web server to understand.
Voice is just one part of the equation. More is required for a complete solution.
Peter
here’s the problem with voice recognition to replace text input in phone applications: i like to work silently. the primary reason for me to use sms to communicate is when i don’t actually want to be speaking. and i want to do email/sms/surf, etc. without making a sound. like when i’m on the bus, waiting for my son’s aikido class to finish, or while sitting on the can…
T9 is actually quite good, the problem is most phones don’t let you flip flop between T9 and something like symbol/alphanumeric(multi-tap) input easily; a smart solution there would make T9 much more useful. it could use tolerance for spelling errors on longer words as well.
typo in your website citation – “building-b.com”.
Hi John,
Thanks for helping fix the error. Appreciate it.
nothing new. this company called ‘promptu’ has a similar app – and they have been around for quite some time i guess.
Dont see this catching on with the mass consumer or enterprise market. Many have tried before… and still trying now.
Sounds very exciting and could be the future for data input on mobile phone devices but I’m a little concerned on how it would handle regional dialects in the UK or any other country for that matter. I struggle to understand people that come from the North East of England for example so how will a software / hardware solution manage? I watch this space with interest.
I think technology is not going to used by most of the consumers.Nothing new and innovative in it.
Sound interesting. Language is a problem as Darren said! Lets see how it goes! Thanks.
I think voice recognition is ok but to a point but i prefere to type what i want to find as insurance to find what i want, as background noise can effect the outcome, ithink, and people may feel silly saying what they want to a phone
It is not a solution…it is a companion. Any reduction in use of the keypad/touchscreen allows for less wear and tear on the components and especially for some phones makes texting a viable option, when otherwise tedious. I’m interested to see if it works on PDAs and other devices.
Does anyone know if this company is still around, if the software works to render notes, if it is available for the N95, and if not, if there is any other software out there that will do voice to text notes for N95?
Thanks.