Audio Analytic Closes $12 Million Funding Round to Extend its Sound Detection Products and IP
Audio Analytic announced yesterday that it had closed a new $12 million round of funding led by Cambridge Innovation Capital and IQ Capital—both existing investors—along with new investor National Grid Partners. With this new financing, Audio Analytic’s total funding to date is nearly $20 million. PitchBook data show that Audio Analytic had previously raised $7.38 million although the company says it is a bit higher, nearing $8 million. The last significant funding round for the company came at the end of 2016 in the amount of $5.5 million.
The company was originally formed in 2008 when CEO and founder Chris Mitchell registered for a number of patents. He then traveled to the U.S. on a Kaufman scholarship “to explore commercial pathways for sound recognition” technology. Audio Analytic then started commercial operations in 2010 according to a company spokesperson.
The Shazam for Everyday Sounds and Audible Anomalies
Technology from Audio Analytic is to sound recognition what the Shazam app is to music recognition or Alexa is to speech recognition. However, instead of music or speech, Audio Analytic’s sound recognition models identify other sounds such as breaking glass, alarms, babies crying, and other everyday noises that may be important to record or be alerted about. The announcement summarizes the technology as “embedded into consumer devices to make them more helpful to people, by understanding and reacting to the contextual information provided by sounds.”
Once a specific sound is recognized, the system can then be used to create alerts or initiate other processes. Audio Analytic’s products include:
- Alexandria™ is a commercially-usable audio dataset for machine learning, featuring millions of audio files that are organized taxonomically, with full data provenance, that the company claims the world’s largest of its kind.
- AuditoryNET™ is a deep neural network for sound recognition, which models the ideophonic features of sounds.
Audio Analytic is embedded in at least three publicly announced products including Hive Hub 360 from Hive, Freebox Delta smart speaker and set-top box from Iliad, and Frontier’s smart speaker. The company says it previously worked with Bragi before it exited the hardware business and has other customers that it cannot reveal publicly due to non-disclosure agreements.
More Investment in Products and IP
“This investment enables us to invest in core technology and IP, further extend the number of sounds and scenes that our customers’ products can recognise. It supports the continued rollout of sound recognition to disrupt a broad range of devices such as smart speakers, hearables, automotive and mobile,” said Chris Mitchell. The company has more than doubled its staff since the early 2017 funding and now has “50 experts” working in the core sound recognition technology. A company spokesperson told Vociebot that all of these are full-time employees and “three quarters of our staff are engineers working across data, platforms, research and embedded software development.” These figures suggest total headcount is around 70 employees.
The need for increased investment may reflect the growing market for sound recognition, but it is also likely driven by increased competition. Sensory announced at CES in January 2019 a new sound recognition product for OEMs called Truly Secure Sound ID and more recently Amazon has rolled out its own solution called Alexa Guard in its line of Echo smart speakers. However, the Amazon competition may be beneficial if sound recognition becomes considered a standard feature for smart speakers and other home-based IoT devices. That would cause more OEMs to seek a partner for sound recognition and Audio Analytic is well positioned to respond to those requests.
Investors Value Independence from Amazon and Google
This announcement is another example of a significant technology company financing round focused on speech and sound recognition that does not rely on Amazon’s or Google’s ecosystem. We have noted previously that companies focused on the Alexa skill and Google Action enablement are typically raising in the single-digit millions for funding rounds while companies independent from the platforms but leveraging similar technologies are routinely raising eight-figure rounds. SoundHound, Clinc, Audioburst, Verbit, Voicea (formerly known as Voicera), and many others follow this pattern as does Audio Analytic.
You can read more about the Audio Analytic announcement here.