Arguably the biggest challenge faced by an IVR system isn’t pronunciation or fast speech recognition. It’s background noise.

In an ideal world, you would always be able to speak into your phone in a quiet place. In the real world, however, there are often other sounds that enter into the equation, whether that’s traffic noise, the wind or chatting coffee shop customers.

Programming IVRs to discriminate between the sound of your voice and that of others in the background is no easy task. So why, we have to ask, does it seem so easy for our brains to do it all on their own?

Think about, for example, the last time you were in a crowded coffee shop with a friend.

You sat down to drink your lattes and, despite all of the other loud conversations going on within a few feet of your table, you’re still able to carry on a conversation.

If you decide that the couple two tables over is more interesting than your friend, you can easily refocus and eavesdrop.

By studying how the human brain makes this switch, we could uncover a new approach to IVRs’ selective listening skills and make a system even better at ignoring background noise and listening just to you.

According to new research from two scientists at the University of California San Francisco, Dr. Nima Mesgarani and Dr. Edward Chang, the reason for our focused attention is a region in our brains that controls auditory analysis.

By emphasizing certain sounds (i.e., the sound of your friend’s voice) and de-emphasizing others (i.e., the couple at the next table down), your brain figures out what’s important and only processes that language.

This naturally occurring cocktail party effect could mean a new approach to speech recognition, which means one day an IVR system that understands you even if you’re dialing in from a crowded restaurant or a concert.