earBuddy is an idea that emerged as a project in "Ubiquitous Computing" class. The concept is to enable communication through noise cancelling headphones by recognizing when the user is being approached and their name is said and reducing the volume of the users music so that they can react to the other person.
The prototype for these headphones was realized by creating a smartphone application, that runs in the background while the user is using the same phone to listen to music. We used a java implementation of CMUSphinx to realize the keyword recognition but other implementations have been considered (Hexagon SDK).
However, this idea can also be realized as a hardware product: Noise Cancelling Headphones with Voice Recognition (using a microphone array).
However, this idea can also be realized as a hardware product: Noise Cancelling Headphones with Voice Recognition (using a microphone array).
Screenshot of the prototype app
Once the name of the user (the keyword) is recognized, the phone loweres the volume of the media volume so that the user can engage in the conversation. If the user does not engage in conversation, the media volume will go back to the normal level. If the user does engage, the music is paused.
