9th International Conference on Software Engineering (SOENG 2023)

June 24 ~ 25, 2023, Copenhagen, Denmark


Accepted Papers


Application of Soft Computing Techniques for Predicting the Efficiency Ofvibration Isolation System

Abir Sarkar1, Ankita Mazumdar2 and Debjit Bhowmik2, 1Department of Civil Engineering, Swami Vivekananda University, Barrakpore, West Bengal, India, 2Department of Civil Engineering, National Institute of Technology, Silchar, Assam, India

ABSTRACT

In today’s world, the application of Artificial Intelligence (AI) has gained attention across various fields. Soft computing techniques, namely Artificial Neural Networks (ANN) and Support Vector Machines (SVM), have shown great potential due to their simplicity and ease of application across multiple fields. However, its application in the field of vibration isolation is unexplored. The screening effectiveness for both open and infilled trenches (OT and IFT) for various trench geometries was assessed in the current study using a two-dimensional finite element (FE) analysis. The study further explores the application of ANN and SVM. The Gaussian Radial Basis Function (GRBF) kernel function for the SVM model modelled the problem better than linear, Exponential Radial Basis Function (ERBF) and polynomial kernel functions. The study suggests that AI models are superior and outperformed traditional MVR models. The AI models were able to achieve coefficient of regression (R2 ) values of 0.99.

KEYWORDS

ANN, SVM, Kernel Functions, MVR, Vibration Isolation.



A Context-aware Intelligent Computer Parental Control and Rewarding Management Application Using Artificial Intelligence and Computer Vision

Darren Huang1 and Yu Sun2, 1Sage Hill School, 20402 Newport Coast Dr, Newport Beach, CA92657, 2Computer Science Department, California State Polytechnic University, Pomona, CA91768

ABSTRACT

In this project, we propose a solution to the problem of childrens exposure to harmful content on their mobiledevices [5]. Our solution involves an application that monitors the content that children view on their devices andalerts parents or guardians if any potentially harmful content is detected. The application utilizes natural languageprocessing and machine learning technologies to analyze the content of messages and images [6][7]. We faced several challenges during the development of our application, including ensuring that the applicationremained easy to use while providing accurate analysis of the content. We also had to ensure that the applicationcould run smoothly on a variety of mobile devices with dif erent operating systems [13]. To test our application, we conducted experiments in various scenarios, such as monitoring childrens device usageduring playtime, study time, and bedtime. Our results showed that our application was highly ef ective in detectingpotentially harmful content, and it was well-received by parents who tested it. Our idea is ultimately something that people should use because it provides an additional layer of protectionforchildren against harmful content. With the increasing prevalence of mobile devices in childrens lives, there is agrowing need for solutions that help parents monitor and regulate their childrens device usage. Our applicationprovides an easy-to-use and ef ective solution to this problem.

KEYWORDS

Parental Control, Management, Artificial Intelligence, Computer Vision


AI for Doctors (Medibot): a Zero Code Solution for Efficient Medical Data Analysis Using Artificial Intelligence

Zachary Liang1, Aleksandr Smolin2, 1Margaret’s Episcopal School, 31641 La Novia Ave, San Juan Capistrano, CA 92675, 2Computer Science Department, California State Polytechnic University, Pomona, CA 91768

ABSTRACT

The efficient analysis of medical data is a critical challenge faced by doctors, who often have the knowledge of what medical information they need to analyze but lack the proper coding knowledge to create a program that can accomplish what they need [1]. This paper presents Medibot, a zero-code solution that enables doctors to use artificial intelligence (AI) to analyze medical data efficiently [2]. Medibot is designed to be intuitive for the user, requiring no prior coding knowledge or experience. The program leverages advanced machine learning algorithms and image analysis techniques to analyze medical data and provide insights and recommendations to the user [3]. Medibot can be used for a variety of medical applications, including the diagnosis of medical conditions and the analysis of patient data. The paper also presents the results of a pilot study conducted with doctors using Medibot, which demonstrated its effectiveness in improving the efficiency and accuracy of medical data analysis. Overall, Medibot represents a promising solution for improving healthcare outcomes using AI and data analytics [4].

KEYWORDS

Medical, Data analysis, Artificial Intelligence.


Integrating Ai With Green Internet of Things in Healthcare for Achieving Un’s Sdgs

Dr. Gasim Alandjani,Computer Science and Engineering Department, Yanbu Industrial College, Yanbu Industrial City, Kingdom of Saudi Arabia

ABSTRACT

The integration of the Green Internet of Things (Green-IoT) and Artificial Intelligence (AI) in healthcare has the potential to revolutionize a wide range of industries, including healthcare wherein Green-IoT-connected medical devices and wearables can collect health data, which can be analyzed by AI algorithms to improve patient outcomes and support better decision-making by healthcare professionals. The convergence of AI and IoT in the field of smart health coupled with machine learning algorithms are enabling new and innovative solutions for healthcare delivery and management. AI algorithms and machine learning techniques can be used to analyze vast amounts of data generated by IoT devices, such as wearable devices, sensors, and smart home health devices, to provide insights into patient health and well-being. This research presents a review of patients’ healthcare services. Particularly, we first give an overview of essential parameters of patients’ healthcare services through Green-IoT-enabled sensor technologies under the use case scenario. We then present a basic architecture for IoT-based healthcare systems considering key requirements in the light of the UN’s Sustainable Development Goals discussing their strengths and weaknesses in the context of the framework for patients’ healthcare services. Finally, we explored various security threats for AI-based architecture and their solutions with a comprehensive methodology to design robust and resilient patients healthcare services system needed in the context of the UN’s sustainable development goals.

KEYWORDS

Artificial Intelligence, Green IoT, Healthcare, UN’s SDGs.



Comparing Yolov5 and Retinanet Object Detection Models for Highway Trash Detection: a Computer Vision Approach to Mitigating Environmental Impact and Promoting Community Health and Safety

Daniel Guo1, Aleksandr Smolin2, 1Marks School, 25 Marlborough Road, Southborough, MA 01772, 2Computer Science Department, California State Polytechnic University, Pomona, CA 91768

ABSTRACT

Computer vision models usually focus on trash detection in nature such as forests, oceans and beaches [1]. However, highway trash is a very common yet neglected problem, dif erent from the trash found in nature, that many communities struggle with. When you drive along the highway, you can often see many concentrations of trash such as drink cans and plastic bags. This paper looks into dif erent computer vision models to best discern the diferent varieties of trash found on the highway roads [2]. We compared two commonly used computer vision models, Yolov5 and Retinanet, to find out which one would best suit the applications of a highway trash detection model [3][4]. These computer vision models were trained on a kaggle dataset of commonly found trash and their categories; we first looked at the dif erent validation steps of the models such as the environment of images and then we tested the two dif erent models on urban images of a variety of trash, allowing us to determine which model was most fit for the discovery of trash on highways. Computer vision models have traditionally been applied to detect trash in natural environments such as forests, oceans, and beaches. However, highway trash is a widespread yet often overlooked problem that many communities struggle with. Unlike natural debris, highway trash is typically composed of drink cans and plastic bags, among other items. In this paper, we investigate the use of computer vision models to identify the dif erent types of trash commonly found on highways. Specifically, we compare two widely used models, Yolov5 and Retinanet, to determine which one is better suited for developing a highway trash detection model [5]. We trained these models on a kaggle dataset that contains various categories of commonly found trash and validated the models under diferent conditions. Subsequently, we tested both models on urban images of a variety of trash to determine their performance in identifying trash on highways. Our results suggest that Yolov5 is better suited for this task.

KEYWORDS

Yolov5, Retinanet, taco trash dataset(Kaggle), Python.


An Multi-step Prediction Algorithm for Analysis of Gravitational Waves Based on Deep Learning

Chunlin Luo1, 2, Yuewei Zhang1, 2 and Jie Zhu1, 2, 1, 2Department of Electronic Engineering, Shanghai Jiao Tong University, 800 Dongchuan Road, Shanghai 200240, China, 1, 2Shanghai Frontiers Science Center for Gravitational Wave Detection, 800 Dongchuan Road, Shanghai 200240, China

ABSTRACT

For the detection of gravitational wave (GW) events, a large number of precise gravitational wave templates are required, but the accuracy and effectiveness of template waveform synthesis still need to be improved. Nowadays, deep learning techniques used for waveform prediction will aid in the creation of template bank across domains. In this paper, we present a multi-step prediction method based on deep learning, which is trained by a large amount of GW waveforms originating from aligned spin binary black hole (BBH) merge, and then the trained model is able to forecast this kind of GW signals and create the GW template bank effectively. The proposed multi-step prediction method effectively utilizes the data conveyed by the waveform sequence and predicts the subsequent waveform sequences directly from the input waveform sequence, avoiding the cumulative error and facilitating easier acquisition of the desired prediction. The experiment results show that for a test set of 100,000 waveforms, the mean value of model predictions and template matching is more than 99.6%. The model also produces a satisfactory prediction when used to forecast the GW waveforms with a wider range of parameters.

KEYWORDS

Gravitational Wave, Waveform Prediction, Multi-step Prediction, Deep Learning.


Counter − Offset Mode: an Alternative Interpretation of Counter Mode With Resistance Against Differential Cryptanalysis

Yasir Nawaz, Shanghai Jiao Tong University

ABSTRACT

The 𝐶𝑜𝑢𝑛𝑡𝑒𝑟 mode is a simple, efficient and widely used mode of operation for a block cipher. It comes with a security proof that guarantees no attacks up to the birthday bound, and have been proven secure against chosen plaintext attack up to2𝑛2⁄encrypted data blocks. We refine the 𝐶𝑜𝑢𝑛𝑡𝑒𝑟 mode with small additional overhead which is known as the 𝐶𝑜𝑢𝑛𝑡𝑒𝑟 − 𝑂𝑓𝑓𝑠𝑒𝑡 mode that is very simple, fully parallelizable and efficient compared to conventional privacy-only 𝐶𝑜𝑢𝑛𝑡𝑒𝑟 mode. The 𝐶𝑜𝑢𝑛𝑡𝑒𝑟 − 𝑂𝑓𝑓𝑠𝑒𝑡 mode take unpredictable input underlying the block cipher this lead to achieve higher resistance against differential cryptanalysis and improve the security of 𝐶𝑜𝑢𝑛𝑡𝑒𝑟 mode without breaking its important advantages. We analyze the concrete security of 𝐶𝑜𝑢𝑛𝑡𝑒𝑟 − 𝑂𝑓𝑓𝑠𝑒𝑡 mode same as provided by 𝑀. 𝐵𝑒𝑙𝑙𝑎𝑟𝑒 𝑒𝑡. 𝑎𝑙. For 𝐶𝑜𝑢𝑛𝑡𝑒𝑟 mode.

KEYWORDS

Modes of operation, symmetric encryption schemes, 𝐶𝑜𝑢𝑛𝑡𝑒𝑟 mode, concrete security, pseudorandom function..