Low Power Tiny Binary Neural Network with improved accuracy in Human Recognition Systems (original) (raw)
Related papers
An Approach of Binary Neural Network Energy-Efficient Implementation
Electronics, 2021
Binarized neural networks (BNNs), which have 1-bit weights and activations, are well suited for FPGA accelerators as their dominant computations are bitwise arithmetic, and the reduction in memory requirements means that all the network parameters can be stored in internal memory. However, the energy efficiency of these accelerators is still restricted by the abundant redundancies in BNNs. This hinders their deployment for applications in smart sensors and tiny devices because these scenarios have tight constraints with respect to energy consumption. To overcome this problem, we propose an approach to implement BNN inference while offering excellent energy efficiency for the accelerators by means of pruning the massive redundant operations while maintaining the original accuracy of the networks. Firstly, inspired by the observation that the convolution processes of two related kernels contain many repeated computations, we first build one formula to clarify the reusing relationships...
A Resource Constrained Neural Network for the Design of Embedded Human Posture Recognition Systems
Applied Sciences
A custom HW design of a Fully Convolutional Neural Network (FCN) is presented in this paper to implement an embeddable Human Posture Recognition (HPR) system capable of very high accuracy both for laying and sitting posture recognition. The FCN exploits a new base-2 quantization scheme for weight and binarized activations to meet the optimal trade-off between low power dissipation, a very reduced set of instantiated physical resources and state-of-the-art accuracy to classify human postures. By using a limited number of pressure sensors only, the optimized HW implementation allows keeping the computation close to the data sources according to the edge computing paradigm and enables the design of embedded HP systems. The FCN can be simply reconfigured to be used for laying and sitting posture recognition. Tested on a public dataset for in-bed posture classification, the proposed FCN obtains a mean accuracy value of 96.77% to recognize 17 different postures, while a small custom datas...
Sensors
The increasing diffusion of tiny wearable devices and, at the same time, the advent of machine learning techniques that can perform sophisticated inference, represent a valuable opportunity for the development of pervasive computing applications. Moreover, pushing inference on edge devices can in principle improve application responsiveness, reduce energy consumption and mitigate privacy and security issues. However, devices with small size and low-power consumption and factor form, like those dedicated to wearable platforms, pose strict computational, memory, and energy requirements which result in challenging issues to be addressed by designers. The main purpose of this study is to empirically explore this trade-off through the characterization of memory usage, energy consumption, and execution time needed by different types of neural networks (namely multilayer and convolutional neural networks) trained for human activity recognition on board of a typical low-power wearable devic...
Human Activity Recognition on Microcontrollers with Quantized and Adaptive Deep Neural Networks
ACM Transactions on Embedded Computing Systems
Human Activity Recognition (HAR) based on inertial data is an increasingly diffused task on embedded devices, from smartphones to ultra low-power sensors. Due to the high computational complexity of deep learning models, most embedded HAR systems are based on simple and not-so-accurate classic machine learning algorithms. This work bridges the gap between on-device HAR and deep learning, proposing a set of efficient one-dimensional Convolutional Neural Networks (CNNs) that can be deployed on general purpose microcontrollers (MCUs). Our CNNs are obtained combining hyper-parameters optimization with sub-byte and mixed-precision quantization, to find good trade-offs between classification results and memory occupation. Moreover, we also leverage adaptive inference as an orthogonal optimization to tune the inference complexity at runtime based on the processed input, hence producing a more flexible HAR system. With experiments on four datasets, and targeting an ultra-low-power RISC-V MC...
A Hardware-Assisted Energy-Efficient Processing Model for Activity Recognition using Wearables
Wearables are being widely utilized in health and wellness applications, primarily due to the recent advances in the sensor and wireless communication, which enhance the promise of wearable systems in providing continuous and real-time monitoring and interventions. Wearables are generally composed of hard-ware/software components for collection, processing, and communication of physiological data. Practical implementation of wearable monitoring in real-life applications is currently limited due to notable obstacles. The wearability and form factor are dominated by the amount of energy needed for sensing, processing and communication. In this paper, we propose an ultra low-power granular decision making architecture, also called screening classifier, which can be viewed as a tiered wake up circuitry, consuming three orders of magnitude less power than the state-of-the-art low-power microcontrollers. This processing model operates based on computationally simple template matching modules, based on coarse to fine grained analysis of the signals with on-demand and gradually increasing of the processing power consumption. Initial template matching rejects signals that are clearly not of interest from the signal processing chain keeping the rest of processing blocks idle. If the signal is likely of interest, the sensitivity and the power of the template matching modules are gradually increased and ultimately the main processing unit is activated. We pose optimization techniques to efficiently split a full template into smaller bins, called mini-templates, and activate only a subset of bins during each classification decision. Our experimental results on real data show that this signal screening model reduces power consumption of the processing architecture by a factor of 70% while the sensitivity of detection remains at least 80%.
Low-Power Embedded System for Gait Classification Using Neural Networks
Journal of Low Power Electronics and Applications
Abnormal foot postures can be measured during the march by plantar pressures in both dynamic and static conditions. These detections may prevent possible injuries to the lower limbs like fractures, ankle sprain or plantar fasciitis. This information can be obtained by an embedded instrumented insole with pressure sensors and a low-power microcontroller. However, these sensors are placed in sparse locations inside the insole, so it is not easy to correlate manually its values with the gait type; that is why a machine learning system is needed. In this work, we analyse the feasibility of integrating a machine learning classifier inside a low-power embedded system in order to obtain information from the user’s gait in real-time and prevent future injuries. Moreover, we analyse the execution times, the power consumption and the model effectiveness. The machine learning classifier is trained using an acquired dataset of 3000+ steps from 6 different users. Results prove that this system p...