It has been an honour to contribute a lecture to Professor Vijay Janapa Reddi’s course on Tiny Machine Learning (TinyML), alongside other industry experts from Google, Qualcomm, Microsoft and more. Indeed, when it comes to online learning, HarvardX – Harvard University’s extension to edX – is one of the top worldwide educational platforms, with a tradition of inviting field-leading experts to teach alongside prominent academics.

The course sits at the intersection of Machine Learning and Embedded IoT Devices. Its topics include:

  • the design and optimisation of TinyML applications and algorithms
  • an introduction to TinyML frameworks, tools and techniques
  • the design of ultra-low-power (milliwatts) systems and other computationally constrained devices.

My lecture entitled “Data Collection Design for Real World TinyML” summarises our many years of insight and expertise into making sound recognition work outside of the lab and in real world applications for consumers. It covers topics such as sound production, audio phenomena and sound variability, and also discusses common machine learning pitfalls and how to avoid these – from the design of data collection, right through to the evaluation methodology. Our approaches, which we share through this lecture, have led us to create ultra-compact embedded sound recognition technology such as ai3-nano™, which we announced in November 2020 and was subsequently pre-validated to run in always-on mode on the Qualcomm Snapdragon 888 Mobile Platform.

The lecture is available to view here for free, alongside other course material and as an add-on to edX’s main TinyML course.

It has been an absolute pleasure collaborating with Harvard’s Professor Reddi and his team and we hope that this lecture will inspire the wider TinyML community to embrace the best practice.

If you want to find out more about our work in the TinyML and MicroML space then check out our blog on ai3-nano™ and read about our challenge in embedding our ai3™ software on the ARM Cortex M0+.


Like this? You can subscribe to our blog and receive an alert every time we publish an announcement, a comment on the industry or something more technical. 


 About Audio Analytic 

Audio Analytic is the pioneer of AI sound recognition technology. The company is on a mission to give machines a compact sense of hearing. This empowers them with the ability to react to the world around us, helping satisfy our entertainment, safety, security, wellbeing, convenience, and communication needs across a huge range of consumer products.

Audio Analytic’s ai3™ and ai3-nano™ sound recognition software enables device manufacturers to equip products at the edge with the ability to recognize and automatically respond to our growing list of sounds and acoustic scenes.