achinoam01: Apple Patent Shows One Use for Nuance's Speech Recognition Technology

Tuesday, May 17, 2011

Apple Patent Shows One Use for Nuance's Speech Recognition Technology

Patently Apple points to a newly published patent application which could show one possible reason Apple is so interested in Nuance's speech recognition technology. Apple is rumored to be in negotiations with Nuance for the use of their speech recognition engine in iOS 5.

The patent application offers a possible solution to two different scenarios where it would be hard for someone to answer a phone in the usual way.

Problem One: Communicating in Noisy Environments
A smartphone user may sometimes have to make or answer a phone call in a noisy environment. Noise could interfere with a phone conversation to a degree that the conversation is no longer intelligible to either conversing party. A user in the noisy environment may try to scream into the phone over the noise, but the screaming and the noise may render the voice signal unintelligible at the other end.
....
Problem Two: Unable to Communicate During a Meeting
In another scenario, it may be inconvenient for a user to talk on a phone. For example, users may be in a meeting and don't want to draw attention to themselves by speaking into the phone. The users may try to whisper into the phone, but the whispering may render the conversation unintelligible. The users may choose to send a text message to the other party, but the other party may be on a landline where texting is unavailable, or may not have a texting plan. It could be frustrating to conduct a telephone conversation when the environment is noisy or the circumstance is inappropriate for a user to speak.

In either scenario, the user is offered the possibility of communicating with the incoming phone call using text-to-speech and speech-to-text capabilities. If receiving a phone call in a loud restaurant or quiet meeting, you might be able to choose the ability to answer using text messaging. The text would later be then turned into speech for the caller to hear, and the caller's verbal responses would likewise be converted back into text.

Now, this is a novel implementation and the likelihood of Apple's implementing it into iOS 5 is probably low. Other perhaps more likely possibilities include the use of voice for search, phone navigation, and dictation for email/sms messages.