Voice Recognition

This week I attended the TFM&A exhibition. Previous years have been hit and miss: two years ago I stayed half a day, last year I was there for 20 minutes.

This year there were some excellent topics in the various theatres and ended up staying a whole day.

The first keynote in the morning was from Google, specifically discussing why mobile search and mobile apps are so important for them.

Basically they find the number of searches from mobile devices so high that their story is that they thought their analytics was wrong due to the growth rates, but alas the data was correct and we are all Googling from our phones a lot.

The big push is on voice recognition. I would recommend everyone to download the Google mobile app because the voice recognition is very good an simple to use.

One problem that came up was during the questions at the end of the Google keynote. A gentleman at the front put up his hand and asked a question. But no one could understand him.

He had a very broad accent and the Google speaker had to keep asking the guy to rephrase he question.

This poses a real problem for speech recognition. After 30 odd years of conversing with people, and understanding the context of what the questions would be about, no humans in the room could understand this one guy! So what chance does a mobile phone have?

3 thoughts on “Voice Recognition

  1. Interesting article. Did you use VR to write it? "He had a very broad accent and the Google speaker had to keep asking the guy to rephrase *he* question."

Leave a Reply

Your email address will not be published. Required fields are marked *