Jibo Tests Sound Localization
Jibo creator Cynthia Breazeal demonstrates sound localization that allows the social robot to recognize where a sound is coming from and turn towards that sound.
Frustration has certainly set in for Jibo’s early backers after the company recently announced delays for the project. Jibo said it “won’t start delivering him in the March/April timeframe as previously communicated,” adding that details of the new delivery date should be announced April 29.
Jibo is even offering refunds for the early backers who don’t want to wait any longer. The company remains confident Jibo will be “worth the wait,” and it has been releasing videos of what the social robot can do, maybe in part to help subdue the growing anger from the backer community.
In the newest video, Jibo creator Cynthia Breazeal demonstrates sound localization that allows Jibo to recognize where a sound is coming from, so that he can turn and orient himself appropriately towards the person speaking. The company said this is one of Jibo’s most useful capabilities, being able to “listen for a speaker and turn to face them, much like a person does when they hear a noise - like an opening door - and turn to face that noise in the room.”
In the video above, you’ll see Breazeal moving around a room and speaking to three Jibos, all of which successfully find where Breazeal’s voice is coming from.
In March 2016, Jibo released a video showing its ability to order you a pizza. This is pretty cool, if it’s truly how Jibo will work, and this wasn’t just a scripted video. Jibo said this application was built by an internal Jibo developer using the Jibo SDK who was playing with an external web service for pizza ordering.
Expect more of these demo videos to come as Jibo tries to keep its backers on board.