Audio Analytic

Audio Analytic is a UK-based company specializing in artificial audio intelligence software, providing consumer electronics with the ability to recognize and respond to sounds beyond speech and music. The company was established in 2008 by Dr. Chris Mitchell, who founded the company based on his PhD research at Anglia Ruskin University into sound information retrieval. After identifying a gap in the market for environmental sound detection, he launched Audio Analytic to pioneer this technology. The company commenced commercial operations in 2010.

The firm’s core business revolves around licensing its proprietary sound recognition software to original equipment manufacturers (OEMs) and chip companies in the smart home and consumer electronics sectors. Its technology is designed to be embedded directly into devices, enabling them to interpret contextual information from their acoustic environment. This allows products like smart speakers, video doorbells, smartphones, and hearables to identify specific sounds such as breaking glass, smoke alarms, a baby crying, or a dog barking and trigger an appropriate action or alert. The business model operates on a cloudless, edge-based AI approach, meaning all processing occurs on the device itself, which is a crucial selling point for privacy and efficiency.

Audio Analytic's main product is its software framework, which has been offered in different versions like ai3™ for a wide range of devices and ai3-nano™ for power-constrained products like true wireless earbuds. This software is delivered as a comprehensive SDK with libraries, reference implementations, and debugging tools. The accuracy of its software is powered by two key internal assets: Alexandria™, one of the world's largest commercially-usable audio datasets for machine learning, and AuditoryNET™, a highly optimized deep neural network for sound recognition. In November 2022, Audio Analytic was acquired by Meta and joined its Reality Labs division to contribute to the development of AR and VR technology by making virtual sounds more realistic.

Keywords: sound recognition, audio intelligence, embedded software, edge AI, consumer electronics, smart home, ai3 software, AuditoryNET, Alexandria dataset, acoustic event detection, machine learning, audio analytics, context-aware devices, IoT, hearables, AR/VR audio, Chris Mitchell, Meta acquisition, audio research, sound classification