Cochlear.ai, a startup that is working towards developing a software that can identify any sort of sound, recently announced to have raised $2 million in a Series A funding round. Cochlear.ai will use these funds to increase the data set of sounds used to train its deep learning algorithms and expand its team by hiring new employees.
The company was founded in Seoul in 2017 by a few colleagues from the Music and Audio Research Group, Seoul National University. Cochlear.ai aims to create a machine listening system that can understand any sort of sound because modern AI systems’ hearing ability is too limited to speech recognition. At the same time, it is only a small part of audio cognition. The company believes that it is the next generation of sound recognition systems. Cochl.Sense, the company’s SaaS can detect about 40 different sounds grouped in three different categories: emergency detection sounds like glass breaking, screaming, sirens, etc.; human interaction like finger snaps, claps, or whistles, etc.; and human sounds like coughing, sneezing, snoring, etc. As the product recognizes these types of sounds, it is quite tricky to train as the developers need to take care of many factors like frequency ranges, sound overlapping, etc. It is available as a cloud API and edge SDK.
The company plans on combining the benefits of Cochl.Sense with devices like smartwatches, smart speakers, vehicles, etc. The startup plans to work with many different industries but is currently focusing on intelligent consumer devices and automotive sectors.
As speech recognition technology is developing fast, many companies are now expanding the speech recognition domain to all types of sounds. Tech giants like Google and Apple are adding context-aware sound recognition to their products. The market for sound recognition technology is growing fast, giving us a glimpse of fully equipped AI sound recognition systems in the future.