Applied and Computational Engineering

- The Open Access Proceedings Series for Conferences

Volume Info.

  • Title

    Proceedings of the 2023 International Conference on Machine Learning and Automation

    Conference Date

    2023-10-18

    Website

    https://2023.confmla.org/

    Notes

     

    ISBN

    978-1-83558-287-9 (Print)

    978-1-83558-288-6 (Online)

    Published Date

    2024-01-31

    Editors

    Mustafa İSTANBULLU, Cukurova University

Articles

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230115

    A comparative study of flexible and rigid hand-oriented exoskeleton robots

    The purpose of this paper is to compare and evaluate the performance differences between flexible and rigid hand exoskeletons in terms of functional recovery and assistance in daily activities.Flexible hand exoskeletons are lightweight and soft devices with stretchable materials and flexible mechanisms designed to mimic the flexibility and versatility of natural hand movements.They typically consist of elastic materials, sensors and actuators that enable natural hand movements and provide light strength support.The main advantages of flexible hand exoskeletons are their comfort and flexibility, and their ability to provide personalized assistance for a variety of daily activities and tasks.This form of design is suitable for patients who require mild hand support and dexterity, such as individuals with mildly impaired hand motor function or who need to perform fine motor movements. In contrast, a rigid hand exoskeleton is a more rigid and stable device that uses robust materials and a rigid mechanism designed to provide a greater degree of strength support and stability.They are typically constructed of metal or composite materials, have a high degree of rigidity and stability, and provide strength support through electrical motors or hydraulic systems.The main advantage of a rigid hand exoskeleton is its higher force output and stability for tasks that require higher loads or complex movements.This form of design is suitable for patients who require greater strength support and stability, such as individuals with reduced hand muscle strength or who need to carry heavy loads

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230116

    Comparative analysis of obstacle avoidance sensors based on assistive intelligent wheel chair

    With the advent of an aging society and the increase in the number of physically disabled people, the pressure faced by medical escorts is gradually increasing. At the same time, since technology is developing rapidly, how to apply wheelchairs to assist the elderly and the disabled has become an urgent problem at this stage. Among them, intelligent wheelchairs with obstacle avoidance function are gradually improving the daily life of the lower limb disability, and the method of obstacle avoidance for intelligent wheelchairs is mainly based on distance measurement technology, and according to the safe distance to determine whether the obstacle affects the security of operator or not, to effectively avoid crashing and falling. The paper introduces six types of obstacle avoidance methods based on different distance measurement sensors: ultrasonic obstacle avoidance, binocular vision obstacle avoidance, structured light obstacle avoidance, Infrared ranging module based on Triangulation and Time of Flight(TOF) method obstacle avoidance, and Light Detection And Ranging(LiDAR) obstacle avoidance. Then it divides them into two categories based on the different principles of distance measurement, collects various parameter information from the official websites of different brands of various sensors, compares their performance parameters, elaborates the working principles of distance measurement, states the advantages as well as limitations of different kinds of sensors, looks forward to their development direction at the end of this paper.

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230117

    Brain-computer interface technology for rehabilitation exoskeleton applications

    In recent years, the development of brain-computer interface (BCI) and exoskeleton technology has received more and more attention. Brain-computer interface, a technology that allows the human brain to communicate with electronic devices or computer programs, has potential applications in sports rehabilitation for the disabled and smart home control. Exoskeleton technology, on the other hand, provides humans with enhanced movement and strength, offering new possibilities for improving the quality of life for people with mobility disorders. Several applications of brain-computer interface and exoskeleton technology are discussed. Applications of brain-computer interfaces range from motor rehabilitation, allowing patients to regain control of paralyzed limbs, to controlling virtual environments and assistive devices. Exoskeletons, on the other hand, enable people with reduced mobility to walk again, providing them with more independence and function. This paper introduces the latest research progress of BCI and exoskeleton technology, including the latest breakthroughs in machine learning and artificial intelligence algorithms, which greatly improve the accuracy and speed of BCI control. Challenges such as developing lightweight and user-friendly exoskeletons and addressing safety and ethical issues in BCI applications are also discussed. The purpose of this paper is to provide a comprehensive and up-to-date reference for researchers and science enthusiasts interested in brain-computer interface and exoskeleton technology.

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230118

    Research on predicting football matches based on handicap data and BPNN

    Football is one of the most influential sports in the world, and billions of people around the globe pay much attention to the football matches. With the growing popularity of football and the continuous development of the football betting industry, the prediction of the outcomes of football matches has become a hot topic in the commercial operations of sports especially footballs in recent years. It is also an important subject of academic research. In this paper, we develop a football match result prediction model based on the back propagation neural network. We take the German Bundesliga competitions as the research object in this paper. In addition to utilizing historical statistic data and team attributes from previous matches, we also incorporate a new dataset, known as handicap data, which refers to the odds data of the football matches, as the input layer of the BPNN (back propagation neural networks) for prediction. We also innovatively use varying numbers of hidden nodes, which greatly improves the prediction accuracy and stability of the model. Experimental results indicate that the average prediction accuracy of this football match prediction model is around 57.2%, with the highest prediction accuracy reaching 59.8% and the lowest prediction accuracy at 53.8%. The prediction model demonstrates relative stability, with no significant fluctuations in prediction accuracy.

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230119

    A review of techniques and methods for deep learning techniques in driver fatigue detection

    Road accidents in which fatigue driving is a significant cause of death are responsible for many deaths worldwide. Approximately 100,000 crashes are caused by driver fatigue each year. Also, fatigue driving is responsible for about 16% of road accidents in general and more than 20% of highway accidents, so fatigue driving accounts for a large percentage of vehicle accidents. Fatigue driving detection usually uses subjective and objective methods. Subjective methods rely on analysing the driver's psychological and facial expression information, while objective methods use external devices to extract feature parameters and apply artificial intelligence algorithms. However, these methods have limitations, such as subjectivity and individual differences. Deep learning, a promising tool inspired by neural networks, offers automatic feature learning, robust pattern recognition, and high adaptability. This review explores the application of deep learning in fatigue driving detection. It examines various deep learning feature extraction methods, classification models, prediction models, and related datasets. By leveraging deep learning techniques, fatigue driving detection can achieve higher accuracy and effectiveness, providing a reliable solution to this critical road safety problem. The review concludes with recommendations and future perspectives in this area.

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230120

    Analysis of the development of an STM32-based smartwatch

    Sensors and microcontrollers are well-functioning and inexpensive, and the smartwatch industry continues to grow. In response to the problems of traditional smartwatches, which are not fully functional, have poor computing effectiveness and are not suitable for wearing, this thesis designs a smartwatch based on the STMicroelectronics NUCLEO-L476RG microcontroller, using Altium Designer 17 software to draw The schematic diagram and Printed Circuit Board are drawn using Altium Designer 17 software, the driver code of the sensor is compiled and integrated using MBED software, the signal is processed by the STMicroelectronics NUCLEO-L476RG microcontroller and displayed in the Organic Light-Emitting Diode with Bluetooth debugger The measured parameters, such as heart rate, ambient temperature, number of steps, latitude and longitude, are displayed in the Organic Light-Emitting Diode with Bluetooth debugger. The results of the study show that the transmission of the signals in the STMicroelectronics NUCLEO-L476RG microcontroller and the display of the signals in the Organic Light-Emitting Diode and Bluetooth debugger have the advantages of small size, perfect functionality and low energy consumption, making it convenient for the user's daily use.

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230121

    Common problems in file conversion and processing parameters of FDM 3D printer

    Fused Deposition Modeling (FDM) 3D printing technology is becoming increasingly popular in manufacturing and rapid prototyping. However, when using FDM 3D printers, one often encounters specific issues. This article summarizes common problems during file conversions and parameter settings with FDM 3D printers. Firstly, this article will discuss issues related to conversion from CAD model to STL file and STL file to Gcode. Secondly, we explore issues related to processing parameter selection and adjustments, including temperature settings, layer height, and print speed. By understanding these problems and their solutions, users can better tackle the challenges encountered using FDM 3D printers and achieve high-quality printing results.

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230122

    Interaction mode enables user perception recognition and perception optimization: An AI human-computer interaction study

    Artificial intelligent (AI) has various ways of human-computer interaction, but most of them overlook the recognition of human perception. If the interaction mode is combined with psychology, the user's mood change can be identified by the user's subtle expression, movement change and voice tone change, so as to provide corresponding services and improve the user experience. Statistical analysis of human responses to different situations in cognitive psychology, incorporating them into human-computer interaction methods. The current human-computer interaction modes in products tend to be standardized, and focusing user experience on user perception will bring special experiences to users. Emotional recognition is a cross disciplinary discipline with broad application prospects, but it has not yet reached a mature stage and requires corpus enrichment, theoretical strengthening, and method innovation. The era of artificial intelligence is leading a new wave of technological progress, and emotion recognition, as an important topic in the field of artificial intelligence, can help computer intelligence recognize human emotions and make human-computer interaction more friendly. In the near future, research on emotion recognition technology will make greater progress and be better applied to practical products.

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230123

    Driver's hazardous state detection in human-computer interaction of automotive cockpits

    Today, the smart car industry is growing rapidly, the functions of the intelligent cockpit based on human-computer interaction are more and more extensive, and the sales volume of intelligent vehicles continues to rise. The incidence of traffic crashes caused by the unsafe state of drivers remains high. The different behavioral states that drivers may emit during driving is a necessary consideration in the design of the intelligent cockpit. This paper takes the driver's state as the starting point to systematically consider the driver's state detection. Summarizing the driver's state detection from four parts: eye state, limb state, facial state, and language state. This paper introduces the development status of the current four types of detection systems, focusing on eye state recognition and limb state recognition. The key driver's characteristic signals are mainly collected by the camera. The driver's state is judged by deep learning, machine learning, and database. This paper is more systematic and comprehensive than the existing literature. Comprehensive consideration of the driver's state contributes to the driver and passengers.

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230124

    Fatigue and distraction warning system for autonomous vehicle drivers in the process of three-level autonomous driving

    Driver fatigue and distraction is the mainest cause of indirect driving accident;with the development of automatic driving technology, self-driving car is more applications; in the process of level 3 automatic driving driver if fatigue and distraction,in time of emergency, can lead to take over not accident in time. Therefore, some researchers have designed a detection and warning system for driver fatigue and distraction in three-level autonomous driving to promote the taking-over process and thus avoid accidents. To better analyze the feasibility and disadvantages of this system, this paper reviews the designs of several current researchers. This paper first examines the System design ideas, design process , and design results in the studied literature, through the literature system design, system evaluation through the comparison of multiple literature system designs, and then combined with the above analysis process. The results of the study show that, In the studied literature,after making a take-over request (TOR) to the driver, the driver is given a tactile, auditory ,and visual multimodal warning; after the system design, the subjects were given feedback according to the people-centered idea, but there are also false alarms and failure to wake up drivers who are too deeply distracted. The multimodal warning system to wake up the tertiary fatigue and distraction in the process of automatic driving, can wake up the driver and promote the takeover process.However, there are detection errors and alarm subsequent insurance measures loopholes, still need subsequent in-depth research and improvement.

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230125

    New perspective on human-computer interaction-based on speech recognition in driving

    With the popularity of smart driving and in-vehicle intelligent systems, speech recognition plays an important role in driving as a key link of human-computer interaction. However, although speech recognition systems have been widely used, their application inhuman-computer interaction still needs to be improved. The purpose of this paper is to examine or review the latest perspectives on speech recognition as part of human-computer interaction in driving. lt provides an in-depth understanding of the potential and future development of speech recognition in driving by exploring, among other things, new directions and features that can be improved. Through a comprehensive analysis of relevant literature and research results, this paper will provide an overview of the current state of speech recognition in human-computer interaction, point out the challenges and limitations that still exist in the driving environment, and focus on new perspectives on speech recognition in driving that may enable more highly accurate command recognition and natural conversational interaction to enhance the driving experience and safety through the use of advanced speech recognition technologies. Explore how technologies such as artificial intelligence and machine learning can be used to drive the development of speech recognition and address current challenges. ln addition, research advances in the areas of driving behavior analysis, emotion recognition, and personalized driving related to speech recognition will be explored. The review and analysis in this paper will provide valuable references and guidance for further research and development of speech recognition-based driving human-computer interaction systems.

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230128

    The effectiveness of PCA and various hyperparameter settings in SVM and KNN for wine quality estimation

    Wine is popular around the world and wine quality evaluation is focused by the wine companies. Wine quality prediction through machine learning is expected to mitigate the waste of time and money of artificial wine quality prediction. Previous researches focused on simple applications and comparisons of the machine learning methods on the wine dataset, but the exploration of optimal parameters of models lacked. Therefore, this research mainly aimed to determine wine quality based on known data by implementing various machine learning models and find the optimal model for predicting the wine quality. For the optimal model, the detailed value of parameter and setting are aimed to be explored. This paper trained five machine learning algorithms and tested them on a wine dataset. The impact of standardization on different machine learning models was tested. Except for decision tree and AdaBoost, standardization is an effective method to improve the performances of different methods. Support vector machine (SVM) with rbf kernel performed best among different SVM classifiers. K-nearest neighbor (KNN) of twenty-five neighborhood points combined with principal component analysis (PCA) of five principal components showed 90.94% accuracy and it is the optimal algorithm.

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230130

    Impact of algorithmic and data structure implementation to game development

    Many game developers tend to build their proprietary game engines with various algorithmic and data structure implementations. This work discusses a game made using the Pygame library that utilizes a maze generation algorithm, collision detection algorithm, and 2d shadow drawing algorithm to analyze how programming implementation could impact the game's representation, optimization, and future development. In this paper, various algorithms and data structures are compared based on their characteristic and performance, which leads to an analysis of how the game would differ from its current state if alternative implementations are used. It also addresses the issues raised by algorithms and data structures based on their interdependence relationship.

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230131

    Drum dance accuracy detector: A CNN-based pose estimation framework for detecting inaccurate movements in Miao drum dance

    As an intangible cultural heritage derived from the Miao culture, drum dance faces the crisis of losing its inheritance. The main cause behind the issue is the lack of teaching resources and distribution of the drum dance. This work is dedicated to resolving this issue with the usage of the recent advances in artificial intelligence. More specifically, the solution of Incorrect Movement Detection (IMD) is invented, where the core technique lies in the CNN-based pose estimation algorithm. IMD enables the real-time detection of the inaccurate dance movements, which realizes the possibility of self-learning of drum dance. The effectiveness of IMD is already validated by its application in a field research project at a Miao village in Xiangxi, Hunan, China. The proposed system may have a significant impact on the preservation and promotion of Miao culture.

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230132

    Prediction on traffic accidents’ severity levels leveraging machine learning-based methods on imbalanced data

    Traffic accidents are a significant problem in many countries, resulting in thousands of injuries and deaths every year. By estimating the severity of traffic accidents, traffic safety together with the crash survival rates could be improved, by taking effective prevention measures at the location where accidents are plentiful and severe. This paper studies the prediction by different classification methods on traffic accident severity levels. The data set used includes 1.6 million traffic accidents recorded in the United Kingdom, ranging from 2000 to 2016. It is a difficult task, since the levels are imbalanced distributed, making it difficult to classify the records accordingly. To tackle this problem, this work compared several classification methods on the task and evaluates their performances from the aspects of time, accuracy, and adaptability on imbalanced data sets. Experiments suggest that among the methods, the decision tree is the most recommended. This paper also provides suggestions for improvements on similar tasks.

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230133

    Studies advanced in face recognition technology based on deep learning

    Face recognition technology has always been a hot research topic in the computer vision community, and has developed rapidly in recent years. Face recognition aims to build a model and predict the face identity information in a given image, which has been widely used in various aspects of social life, such as identity authentication, security encryption, human-computer interaction, etc. In order to improve the accuracy and speed of face recognition, and how to maintain good face recognition under the premise of occlusion, many advanced technologies have been proposed. This paper summarizes the face recognition technologies proposed in recent years, and introduces the latest research progress in the field of face recognition from two aspects: traditional face recognition based on manual features and face recognition based on deep learning. Specifically, we first briefly introduce traditional face recognition methods. Second, we introduce the mechanism of traditional Convolutional Neural Networks(Hereinafter referred to as CNN) in face recognition. Finally, we focus on the application of Transformer in the field of face recognition. According to the datasets used by the methods introduced above, the performance of these methods is summarized, the advantages and disadvantages of CNN and Transformer are pointed out, and the future development direction is proposed.

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230136

    Pedestrian detection and gender recognition utilizing YOLO and CNN algorithms

    As crowd-based activities continue to surge in locales such as markets and restaurants, the significance of understanding pedestrian flow is increasingly evident. Over recent years, advancements in dynamic pedestrian detection, facilitated by the YOLO (You Only Look Once) algorithm, have seen widespread application in areas like crowd management and occupancy estimation. The YOLO algorithm has demonstrated high accuracy and efficiency in real-time object tracking and counting. However, for specific use cases, data derived solely from monitoring pedestrian flows may prove inadequate. This study presents YOLO-Gender, a system leveraging YOLO and Convolutional Neural Network (CNN) for pedestrian tracking and gender classification. The objective is to enhance the richness of data extracted from surveillance camera footage, thus rendering it more valuable for societal applications. The YOLO suite of algorithms, hailed for their superior performance and rapid iteration speed, is among the most extensively utilized tools in the field. The proposed system is predicated on YOLO v8, the most advanced iteration of the YOLO algorithm, released in 2023, which boasts its highest accuracy to date.

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230137

    Designing a dual-camera highway monitoring system based on high spatiotemporal resolution using neural networks

    The criticality of infrastructure to societal development has seen highways evolve into an essential component of this ecosystem. Within this, the camera system has assumed significant importance due to the necessity for monitoring, evidence collection, and danger detection. However, the current standard of using high frame rate and high-resolution (HSR-HFR) cameras presents substantial costs associated with installation and data storage. This project, therefore, proposes a solution in the form of a High Spatiotemporal Resolution process applied to dual-camera videos. After evaluating state-of-the-art methodologies, this project develops a dual-camera system designed to merge frames from a high-resolution, low frame rate (HSR-LFR) camera with a high frame rate, low-resolution (LSR-HFR) camera. The result is a high-resolution, high frame rate video that effectively optimizes costs. The system pre-processes data using frame extraction and a histogram equalization method, followed by video processing with a neural network. Further refinement of the footage is performed via color adjustment and sharpening prior to a specific application, which in this case is license plate recognition. The system employs YOLOv5 in conjunction with LPRNet for license plate recognition. The resulting outputs demonstrate significant improvement in both clarity and accuracy, providing a more cost-effective solution for highway monitoring systems.

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230138

    Optimizing road sign detection using the segment anything model for background pixel exclusion

    Road sign recognition plays a great role in automatic driving. At present, for the task of road sign recognition, the structure related to CNN is mainly used to classify the detected road signs. However, due to the complexity of road conditions and weather, the detected images often contain complex background pixels, which affects the accuracy of classification. Due to the release of segment anything model, there is a powerful tool for eliminating background pixels quickly and conveniently. We optimize the data to retain only the road sign part of the images, and directly let the model train the image without back- ground pixels to realize the recognition task. This paper constructs two CNN classification models with the same structure through comparative experiments. One model uses the segmented data set for training, and the other model uses the original data set for training. The performance of the two models is evaluated and compared to verify the optimization effect brought by the segmented data set. It is found that segment anything model can accurately cut most of the images in the data set. And the performance of the model trained by the segmented data set is better than that of the model using the original data set. Moreover, the optimized model can achieve high accuracy in less training times.

  • Open Access | Article 2024-01-31 Doi: 10.54254/2755-2721/31/20230139

    Study on the safety of elbow joint with simple upper limb rehabilitation exoskeleton

    Due to the great demand for treating physicians for patients with disabilities caused by stroke, in order to reduce the workload of doctors, exoskeleton robots have been effectively applied and promoted in post-illness health care. Because the human rehabilitation movement needs certain accuracy and low error tolerance rate, the rehabilitation exoskeleton robot requires high precision. Since the exoskeleton robot needs to be worn on the human body, there are certain requirements for the size and wearing comfort of the machine, and the functional structure of all aspects of the upper limb exoskeleton rehabilitation robot still needs to be further improved. In the medical rehabilitation exoskeleton, for the elbow exoskeleton of the upper limb, an elbow exoskeleton with binding structure is proposed in this paper. The elastic structure connects the human body with the binding structure, and on this basis, the wearing comfort is improved through the improvement of the structure and the selection of materials. At the same time, the passive joint between the binding structure and the exoskeleton is designed. In case of motion error when the motor is actively driven, the passive joint reduces the harmful force and torque, so as to ensure safety. Finally, the feasibility of this method is demonstrated through simulation and analysis of the movement of exoskeleton machine. It is proved that the design has certain practicability and reference value for practical treatment.

Copyright © 2023 EWA Publishing. Unless Otherwise Stated