scholarly journals Memory-Efficient AI Algorithm for Infant Sleeping Death Syndrome Detection in Smart Buildings

AI ◽  
2021 ◽  
Vol 2 (4) ◽  
pp. 705-719
Author(s):  
Qian Huang ◽  
Chenghung Hsieh ◽  
Jiaen Hsieh ◽  
Chunchen Liu
Keyword(s):  
Healthcare Services ◽  
Night Vision ◽  
Raspberry Pi ◽  
Sudden Infant ◽  
Detection Accuracy ◽  
Power Budget ◽  
Smart Buildings ◽  
Memory Space ◽  
Memory Footprint ◽  
Memory Efficient

Artificial intelligence (AI) is fundamentally transforming smart buildings by increasing energy efficiency and operational productivity, improving life experience, and providing better healthcare services. Sudden Infant Death Syndrome (SIDS) is an unexpected and unexplained death of infants under one year old. Previous research reports that sleeping on the back can significantly reduce the risk of SIDS. Existing sensor-based wearable or touchable monitors have serious drawbacks such as inconvenience and false alarm, so they are not attractive in monitoring infant sleeping postures. Several recent studies use a camera, portable electronics, and AI algorithm to monitor the sleep postures of infants. However, there are two major bottlenecks that prevent AI from detecting potential baby sleeping hazards in smart buildings. In order to overcome these bottlenecks, in this work, we create a complete dataset containing 10,240 day and night vision samples, and use post-training weight quantization to solve the huge memory demand problem. Experimental results verify the effectiveness and benefits of our proposed idea. Compared with the state-of-the-art AI algorithms in the literature, the proposed method reduces memory footprint by at least 89%, while achieving a similar high detection accuracy of about 90%. Our proposed AI algorithm only requires 6.4 MB of memory space, while other existing AI algorithms for sleep posture detection require 58.2 MB to 275 MB of memory space. This comparison shows that the memory is reduced by at least 9 times without sacrificing the detection accuracy. Therefore, our proposed memory-efficient AI algorithm has great potential to be deployed and to run on edge devices, such as micro-controllers and Raspberry Pi, which have low memory footprint, limited power budget, and constrained computing resources.

Implementasi Metode Unsupervised Learning Pada Sistem Keamanan Dengan Optimalisasi Penyimpanan Kamera IP

2021 ◽  
Vol 5 (6) ◽  
pp. 1099-1105
Author(s):  
Desta Yolanda ◽  
Mohammad Hafiz Hersyah ◽  
Eno Marozi
Keyword(s):  
Unsupervised Learning ◽  
Night Vision ◽  
Raspberry Pi ◽  
Monitoring Systems ◽  
Test Results ◽  
Storage Process ◽  
File Size ◽  
Web Based ◽  
Security Monitoring ◽  
Human Faces

Security monitoring systems using face recognition can be applied to CCTV or IP cameras. This is intended to improve the security system and make it easier for users to track criminals is theft. The experiment was carried out by detecting human faces for 24 hours using different cameras, namely an HD camera that was active during the day and a Night Vision camera that was active at night. The application of Unsupervised Learning method with the concept of an image cluster, aims to distinguish the faces of known or unknown people according to the dataset built in the Raspberry Pi 4. The user interface media of this system is a web-based application built with Python Flask and Python MySQL. This application can be accessed using the domain provided by the IP Forwarding device which can be accessed anywhere. According to the test results on optimization of storage, the system is able to save files only when a face is detected with an average file size of ± 2.28 MB for 1x24 hours of streaming. So that this storage process becomes more efficient and economical compared to the storage process for CCTV or IP cameras in general.

scholarly journals Pendeteksian Septoria pada Tanaman Tomat dengan Metode Deep Learning berbasis Raspberry Pi

2021 ◽  
Vol 5 (1) ◽  
pp. 107-113
Author(s):  
Kahlil Muchtar ◽  
Chairuman ◽  
Yudha Nurdin ◽  
Afdhal Afdhal
Keyword(s):  
Deep Learning ◽  
Leaf Spot ◽  
Raspberry Pi ◽  
Leaf Spot Disease ◽  
Detection Accuracy ◽  
Crop Failure ◽  
Tomato Production ◽  
Spot Disease ◽  
Average Accuracy ◽  
Septoria Leaf Spot

much needed to meet the needs of both industry and households. However, tomato plants still require serious handling in increasing the yields. Data from the Central Bureau of Statistics shows that the number of tomatoes produced is not in accordance with a large number of market demands, resulting from the decrease of tomato yields. One of the obstacles in increasing tomato production is that the crops are attacked by septoria leaf spot disease due to the fungus or the fungus Septoria Lycopersici Speg. Most farmers have limited knowledge of the early symptoms, which are not obvious, and also facing difficulty in detecting this disease earlier. The problem has been causing disadvantages such as crop failure or plant death. Based on this problem, a study will be conducted with the aim of designing a tool that can be used to detect septoria leaf spot disease based on deep learning using the Convolutional Neural Network (ConvNets or CNN) model, where an algorithm that resembles human nerves is one of the supervised learning and widely used for solving linear and non-linear problems. In addition, the researcher used the Raspberry Pi as a microcontroller and used the Intel Movidius Neural Computing Stick (NCS) which functions to speed up the computing process so that the detection process is easier because of its portable, fast and accurate nature. The average accuracy rate is 95.89% with detection accuracy between 84.22% to 100%.  

scholarly journals A Research Paper on Traffic Sign Recognition with Machine Learning and IOT

2021 ◽  
Vol 9 (VI) ◽  
pp. 1774-1778
Author(s):  
Mahesh Singh
Keyword(s):  
Machine Learning ◽  
Object Detection ◽  
Real World ◽  
Research Paper ◽  
Raspberry Pi ◽  
Detection Accuracy ◽  
Great Success ◽  
The Real ◽  
Road Traffic Safety ◽  
Execution Speed

This paper will help to bring out some amazing findings about autonomous prediction and performing action by establishing a connection between the real world with machine learning and Internet Of thing. The purpose of this research paper is to perform our machine to analyze different signs in the real world and act accordingly. We have explored and found detection of several features in our model which helped us to establish a better interaction of our model with the surroundings. Our algorithms give very optimized predictions performing the right action .Nowadays, autonomous vehicles are a great area of research where we can make it more optimized and more multi - performing .This paper contributes to a huge survey of varied object detection and feature extraction techniques. At the moment, there are loads of object classification and recognition techniques and algorithms found and developed around the world. TSD research is of great significance for improving road traffic safety. In recent years, CNN (Convolutional Neural Networks) have achieved great success in object detection tasks. It shows better accuracy or faster execution speed than traditional methods. However, the execution speed and the detection accuracy of the existing CNN methods cannot be obtained at the same time. What's more, the hardware requirements are also higher than before, resulting in a larger detection cost. In order to solve these problems, this paper proposes an improved algorithm based on convolutional model A classic robot which uses this algorithm which is installed through raspberry pi and performs dedicated action.

scholarly journals Evaluation of In Vivo Spike Detection Algorithms for Implantable MTA Brain—Silicon Interfaces

2020 ◽  
Vol 10 (3) ◽  
pp. 26
Author(s):  
Mattia Tambaro ◽  
Elia Arturo Vallicelli ◽  
Gerardo Saggese ◽  
Antonio Strollo ◽  
Andrea Baschirotto ◽  
...  
Keyword(s):  
Energy Operator ◽  
Oxide Semiconductor ◽  
Detection Accuracy ◽  
Noise Power ◽  
Spike Detection ◽  
Power Budget ◽  
Detection Algorithms ◽  
Implanted Devices

This work presents a comparison between different neural spike algorithms to find the optimum for in vivo implanted EOSFET (electrolyte–oxide-semiconductor field effect transistor) sensors. EOSFET arrays are planar sensors capable of sensing the electrical activity of nearby neuron populations in both in vitro cultures and in vivo experiments. They are characterized by a high cell-like resolution and low invasiveness compared to probes with passive electrodes, but exhibit a higher noise power that requires ad hoc spike detection algorithms to detect relevant biological activity. Algorithms for implanted devices require good detection accuracy performance and low power consumption due to the limited power budget of implanted devices. A figure of merit (FoM) based on accuracy and resource consumption is presented and used to compare different algorithms present in the literature, such as the smoothed nonlinear energy operator and correlation-based algorithms. A multi transistor array (MTA) sensor of 7 honeycomb pixels of a 30 μm2 area is simulated, generating a signal with Neurocube. This signal is then used to validate the algorithms’ performances. The results allow us to numerically determine which is the most efficient algorithm in the case of power constraint in implantable devices and to characterize its performance in terms of accuracy and resource usage.

An Embedded FTP Server: Research and Implementation

2014 ◽  
Vol 543-547 ◽  
pp. 1977-1980
Author(s):  
Min Gong ◽  
Wei Dong ◽  
Zhen Ya Zhang
Keyword(s):  
User Interface ◽  
Embedded Systems ◽  
Embedded System ◽  
Transmission Protocol ◽  
Graphic User Interface ◽  
Command Line ◽  
Memory Space ◽  
Complicating Factors ◽  
Design Requirements ◽  
Memory Efficient

Traditional File Transmission Protocol (FTP) servers are not suitable for embedded environments since these servers take complicating factors into account and consequently take a lot of memory space. Moreover, some FTP server applications even manage all file transmission by utilizing graphic user interface. However, such server is not adequate for embedded environments. When embedded system needs is more memory-efficient efficient in a resource restricted system. To cope with this problem, in this problem we conduct research on lightweight FTP server for embedded systems. And, we first summarize features of a lightweight FTP server. Then, based on the design requirements for FTP server for embedded systems, a lightweight FTP server is designed. The server adopts a command-line user interface and other advanced features, such as transmission encryption and graphic user interface, are abandoned for efficiency reasons. Last, to evaluate our design of the FTP server, several experiments are conducted and all results show that the FTP server works quite well under embedded environments. With more and more embedded systems emerging in the industry, it can be expected that the software introduced in this paper will play a more and more important role in the industry.

Memory Efficient One-Sided Communucation Library "ACP" in Globary Memory on Raspberry Pi 2

2016 ◽  
Author(s):  
Yoshiyuki Morie ◽  
Hiroaki Honda ◽  
Takeshi Nanri ◽  
Taizo Kobayashi ◽  
Hidetomo Shibamura ◽  
...  
Keyword(s):  
Raspberry Pi ◽  
Memory Efficient

scholarly journals Binocular intelligent following robot based on YOLO-LITE

2021 ◽  
Vol 336 ◽  
pp. 03002
Author(s):  
Yuanyuan Zheng ◽  
Jun Ge
Keyword(s):  
Neural Network ◽  
Real Time ◽  
Deep Neural Network ◽  
Frame Rate ◽  
Raspberry Pi ◽  
Detection Accuracy ◽  
Location Error ◽  
Embedded Devices ◽  
Robot System ◽  
Time Performance

In order to solve the problem that the deep neural network model is large in scale, the calculation time is too long, and the real-time performance is severely limited when combined with embedded devices, so studied the intelligent follower robot system based on YOLO-LITE algorithm combined with Raspberry Pi 3B+. The system mainly includes camera processing, target detection and other modules. Obtained the internal and external parameters of the camera through calibration, and according to these parameters to correct the binocular camera. Recognized and located the target in each frame of image, calculated the distance from the camera to the target and the center location error, and driven the car to move. The experimental results show that the following car has excellent real-time performance, the average detection frame rate can reach 20Fps, and the average detection accuracy can reach more than 80%.

scholarly journals Development of a portable community video surveillance system

2019 ◽  
Vol 9 (3) ◽  
pp. 1814
Author(s):  
S. Fakhar A. G ◽  
A. Fauzan K ◽  
M. Saad H ◽  
R. Affendi H ◽  
K. H. Fen
Keyword(s):  
Video Surveillance ◽  
Surveillance System ◽  
Night Vision ◽  
Raspberry Pi ◽  
Video Surveillance System ◽  
Image Subtraction ◽  
Whole Process ◽  
Video Feed ◽  
Group Members ◽  
Vision Simulation

In 2016, a crime rate has been evidently increasing particularly in Kuala Lumpur areas, including reports on house break-ins, car thefts, motorcycle thefts and robbery. One way of deterring such cases is by installing CCTV monitoring system in premises such as houses or shops, but this usually requires expensive equipment and installation fees. In this paper a cheaper alternative of a portable community video surveillance system running on Raspberry Pi 3 utilizing OpenCV is presented. The system will detect motion based on image subtraction algorithm and immediately inform users when intruders are detected by sending a live video feed to a Telegram group chat, as well as sound the buzzer alarm on the Raspberry Pi. Additionally, any Telegram group members can request images and recorded videos from the system at any time by sending a get request in Telegram which will be handled by Telegram Bot. This system uses the Pi NoIR camera module as the image acquisition device equipped with a 36 LED infrared illuminator for night vision capability. In addition to the Python language, OpenCV, a computer vision simulation from Intel is also used for image processing tasks. The performance analysis of the completed system is also presented computational complexity while offering improved flexibility. The performance time is also presented, where the whole process is run with a noticeable 3 seconds delay in getting the final output.

Sign in / Sign up

Export Citation Format

Share Document