An Overview of Machine Learning within Embedded and Mobile Devices–Optimizations and Applications

Abstract

Embedded systems technology is undergoing a phase of transformation owing to the novel advancements in computer architecture and the breakthroughs in machine learning applications. The areas of applications of embedded machine learning (EML) include accurate computer vision schemes, reliable speech recognition, innovative healthcare, robotics, and more. However, there exists a critical drawback in the efficient implementation of ML algorithms targeting embedded applications. Machine learning algorithms are generally computationally and memory intensive, making them unsuitable for resource-constrained environments such as embedded and mobile devices. In order to efficiently implement these compute and memory-intensive algorithms within the embedded and mobile computing space, innovative optimization techniques are required at the algorithm and hardware levels. To this end, this survey aims at exploring current research trends within this circumference. First, we present a brief overview of compute intensive machine learning algorithms such as hidden Markov models (HMM), k-nearest neighbors (k-NNs), support vector machines (SVMs), Gaussian mixture models (GMMs), and deep neural networks (DNNs). Furthermore, we consider different optimization techniques currently adopted to squeeze these computational and memory-intensive algorithms within resource-limited embedded and mobile environments. Additionally, we discuss the implementation of these algorithms in microcontroller units, mobile devices, and hardware accelerators. Conclusively, we give a comprehensive overview of key application areas of EML technology, point out key research directions and highlight key take-away lessons for future research exploration in the embedded machine learning domain.

Keeping surgeons in the loop: are handheld robotics the best path towards more autonomous actions? (A comparison of complete vs. handheld robotic hepatectomy for colorectal liver metastases)

“…Automated venous puncture devices using robotics, ultrasound, and computer vision (CV) have been developed that show similar success rates as with humans [43] . In addition to ultrasound, acoustic waves have also been harnessed to give MI surgeons the ability to perceive different tissues, and they may be utilized to permit improved surgical dissection [44] . Recent exponential advances in machine learning, deep learning, and CV make this field of robotics particularly interesting for dissection and again raise concerns that our obsession with haptics may be ill-advised and that many other avenues of AI should also be explored [45] .…”

State of charge estimation of an electric vehicle's battery using tiny neural network embedded on small microcontroller units

“…To the authors' best knowledge, the TensorFlow Lite micro has never been used for the quantization of DL networks dedicated for battery SOC estimation onboard EVs. The deployment of DL on small MCUs has several advantages such as privacy protection (no need for a connected cloud), reduced latency and power consumption, 41 as well as a small memory footprint 42 . This is been verified in Reference 43 where the authors investigated the performance of decision tree (DT), k‐nearest neighbors (KNN), and SVM implemented on an STM32 Nucleo board.…”

Applications of Artificial Intelligence in Myopia: Current and Future Directions

“…Its definition simulates human intelligence through machines ( 20 ). Machine learning (ML) is a branch of AI and mainly uses computer system programming to perform tasks or predict results ( 21 ). ML has great potential in clinical practice and machine translation ( 22 ).…”