This Cambridge-based start-up is training machines to recognise a wide taxonomy of sounds, such as breaking glass, barking dogs and snoring. In the past few years, consumer technology has become very good at understanding the world by sight, but – apart from speech recognition – highly accurate sound recognition has been given little attention.
HOW THEY’RE DISRUPTING
A wide range of consumer technology is being taught to pick out sounds that are important to humans, like smoke alarms, laughter, crying babies and police sirens, for example. This intelligence will then be embedded into smart home systems, smart speakers, mobile phones, hearables and cars that will alert you when an intruder breaks in, tag your videos if somebody laughs or alert you to oncoming and unseen emergency vehicles. The technology can even empower devices to react independently, such as turning on the lights and playing music if a window is smashed; or increasing the distance between you and the car in front if it thinks you may be distracted while driving.
The potential for progress in the world of artificial audio intelligence is massive. Analyst firm IDC estimates that sound recognition software market will be worth $1.2bn by 2021. It is fast becoming vital for smart home technology to detect and interpret a wide range of sounds, so people or devices can respond quickly and easily to what’s going on in their homes when they’re away. Founder Chris Mitchell said: “This is not something that has been tackled before in any meaningful sense, and it’s a huge undertaking.”
INVESTMENTS AND FUTURE
The company has received $8.5m in funding – the most recent round, 2017’s Series A, raised $5.5m. Audio Analytic investors include Cambridge Innovation Capital PLC and IQ Capital Partners LLP.