The Internet of Things (IoT) refers to a system of interrelated, Internet-connected objects, «things,» that can collect and transfer data without human intervention.
The personal or business possibilities are endless. A ‘thing’ can refer to a connected medical device, a biochip transponder (think livestock), or any object outfitted with sensors that can gather and transfer data over a network.
Machine Learning is typically associated with computationally heavy cloud-based solutions with relatively high latencies, high power consumption, and the need for high bandwidths links.
With TinyML, IoT and ML come together by shrinking deep learning networks to fit on tiny hardware, thus requiring low energy, low bandwidth links, and reduced latencies. Moreover, it allows providing privacy to many novel applications. Audio analytics, pattern recognition, and voice human-machine interfaces are the fields where most of TinyML is applied today, targeting applications like children and elderly care, safety, sentiment analysis, and predictive monitoring, among many others.
In this talk, I will first briefly overview IoT with its essential components. Then, I will present TinyML with the current status of the available hardware, software, and some details of the existing applications.
