INTI Institutional Repository
Not a member yet
2178 research outputs found
Sort by
Study on Image Background Removal using Deep Learning
No full textRemoving image backgrounds is a common job in image processing and computer vision. By
isolating the main object from the back, background removal in photographs aims to make it easier
to examine or edit the image. There are numerous methods for removing the background from an
image, including deep learning, color-based segmentation, and human selection. The U-Net
architecture, one of the deep learning-based techniques, has demonstrated encouraging results in
image segmentation tasks, including image background removal. A convolutional neural network
created for biological image segmentation is known as the U-Net architecture. The design consists
of an encoder network that stores the context and a decoder network that generates the
segmentation map. The U-shape of the U-Net architecture enables it to record both the overall
context and the local specifics of the image. For several picture segmentation tasks, including
image background removal, U-Net architecture has undergone modification. The suggested
method for removing image backgrounds using U-Net entails training a U-Net model on a dataset
of pictures with and without background. Then, using the demonstrated methodology, the
backdrop is removed from recent photographs. The suggested method differs from current
approaches in various, including its high accuracy and capacity to handle complicated backgrounds.
Computer vision, object identification, and photo manipulation are just a few of the uses for the
suggested metho
Online Product Evaluation System Based on Ratings and Review
No full textThe decision-making process for product design and improvement is hindered by traditional user
research methods due to the rapid updating pace, limited survey scopes, small sample sizes, and
labor-intensive procedures. This study suggests a novel method for gathering valuable online
evaluations from e-commerce platforms, develops a system for measuring the effectiveness of a
product and suggests ways to improve a product using sentiment analysis and opinion mining of
online reviews. The method's efficacy is supported by a sizable body of user reviews for
smartphones, from which we can reliably estimate the product's unfavorable review rate with only
a 9.9% error using the assessment indication system. After considering the entire method in the
case study, improvement strategies are suggested. The strategy is applicable for product evaluation
Cardiovascular Diseases Detection Using Photo Plethysmography (PPG) Signal Data
No full textPhotoplethysmography (PPG) signals have been widely used in clinical practice as diagnostic
tools. In this article, techniques of machine learning have been used to improve the detection
of cardiovascular disease (CVD) from the PPG signal data. Hypertension and stress are the
main causes of the increase in blood pressure (BP), which in turn causes cardiovascular
diseases. The treatment of patients, mainly those who have been suffering from CVD, resulted
in an increment in the death rate. PPG is non-invasive, low-cost, fast, and simple to use. The
signals of PPG are used for figuring out the anomalies in the cardiovascular system. By using
PPG technology, cardiovascular parameters like blood pressure and heart rate are detected. This
article investigates a machine learning and Deep Learning technique, which is Neural Network
(NN), that has been used to assist physicians, this has achieved an accuracy of 98% by using
the PPG-BP data set
Enhancing Classification Algorithms with Metaheuristic Technique
No full textClassification is a process of grouping or placing data into appropriate categories or classes based on specificattributes or features to predict labels or classes of new data based on patternsobserved from previously trained data. Implementing this process uses classification algorithms such asNaïve Bayes, Support Vector Machine,and Random Forest. However, the classification algorithm cannotclassify data optimally due to the challenges in dealing with variousdata sets. Not all available featureswillmake a solidcontribution to the label of the data class, often in the form of noise or interference. For this reason, it is necessary to carry out a feature selection process. Currently, many feature selection processes have been carried out using correlation values from chi-square and gain-information, but the accuracy of the resultsis often still not good enough. This is because the chi-square and gain-information values are fixed. So,the selection of features is minimaland is not based on the previous learning process or what is known as heuristics. For this reason, in this research,several auxiliary algorithms are introduced to improve the performance of the classification algorithm, namely the meta-heuristic algorithm. Meta-heuristic algorithms are search techniques used to solve complexoptimization problems, and these algorithms can help provide reasonable solutions in a shorter time thanexact methods. In its operation, the metaheuristic algorithm optimizes the feature selection process,which will later be processed using the classification algorithm.Three (3) meta-heuristics were implemented, namely Genetic Algorithm, Particle Swarm Optimization, and Cuckoo Search Algorithm; the experiment was conducted, and the results were collected and analyzed. The result shows that combining Naive Bayes and Genetic Algorithmgives the best performance regarding higher accuracy improvementat +23.77%
Creating a "Ready-to-Use" AI Agent for Navigating Digital Platform to Enhance Collaborative Efficiency
No full textThe study aims to address the prevalent issue of navigation inefficiency within digital collaboration platforms, a critical factor affecting user productivity and satisfaction. This paper identifies research gaps and highlights the potential of multimodal AI Agents in revolutionising user experience in these platforms. Specifically, the development of a ready-to-use AI Agent was designed for the DingTalk platform, capable of guiding users through various digital functionalities via conversational interfaces. By leveraging recent advancements in large language models (LLMs) and the concept of "Model as a Service" (MaaS), the proposed AI solution seeks to overcome current navigation obstacles, thereby enhancing the sustainability and effectiveness of digital work ecosystems
Analyzing Human-Centric Wireless Energy Harvesting for Sustainable and Resilient Energy Systems
No full textDiscussions around wireless communications and energy harvesting in the literature are elitist
and techno-centric. There is a need for more studies on a human-centric approach that
simplifies the concept and links energy harvesting with human empowerment and well-being.
Also, studies so far on the implementation of wireless energy harvesting in rural settlements
have been scanty despite the important role wireless communication plays in social interactions
and financial transactions within rural communities. The challenge of little awareness of the
human sustainability impact of WEH is impeding stakeholders’ advocacy and advancement of
wireless energy harvesting (WEH) despite its real and perceived benefits. So far, research
efforts as reported in the literature are less than proportionate to the potential of WEH. Apart
from providing sustainable energy for wireless devices and networks, its generation of nonelectric
renewable energy is a huge plus for ensuring access to affordable, reliable, sustainable,
and modern energy for all as contained in the United Nations Sustainable Development Goal
7 (SDG 7). Given the astronomical growth in the use of wireless devices globally, more
research studies are needed to create more awareness for massive WEH advocacy. This study
uses a model-driven approach to simplify and highlight critical aspects of WEH such as
wireless body area network, sources of non-electric renewable energy sources, green energy
(environmental protection), intermittent nature of renewable wireless energy sources, and
research directions. Also, the application of WEH for the sustenance of lives and livelihood in
rural economies is examined using the rural community of Adum-Aiona in Nigeria as a case
study. It is expected that the information provided will create more awareness and foster further
discussions around the human-centric angle of WEH even as advances in technology continue
to be made
ESG Indices, Sustainable Practices and Strategic Competitiveness within the Modern Business Landscape
No full textThis paper undertakes the arduous task of examining the accuracy and relevance of a statement
lacking clear benchmarks. The assertion posits that Environmental, Social and Governance
(ESG) indices is indispensable not just for assessing a company’s sustainability efforts, but also
enhancing its competitive edge. Through analysis of existing literature, frameworks, and
empirical evidence, this article elucidates the nuances inherent in the link between ESG indices,
sustainable practices, and strategic competitiveness within the modern business landscape. To
accomplish this, the paper adopts a structured approach, dissecting the statement into its
constituent components to discern the key points concerning ESG. It intends to challenge the
notion of ESG as an indispensable tool in measuring sustainability, while also evaluating
relevant initiatives and potential impacts. Furthermore, the article explores the extent to which
indices truly function as significant tools, addressing societal and environmental issues while
offering pathways for corporate success beyond conventional business performance metrics
Assessment on the Effect of Rapid Urbanization on Groundwater Quality of Mirpur Area, Dhaka City, Bangladesh
No full textDhaka’s population is increasing day by day as more people are migrating towards Dhaka city for improving their level of living and for availing the cities amenities. Hence, this is creating pressure on the overall water quality of Dhaka city. The work intended to assess the groundwater standard in Mirpur area of Dhaka City, Bangladesh by analyzing eight (08) different physicochemical water quality parameters. In December 2023, six (06) separate water samples were taken throughout the dry season from six (06) sites across the area. The results showed that all sites had pH levels between 7.18 to 7.50, which is within acceptable ranges for drinking water quality and indicates neutral to slightly alkaline conditions. Total dissolved solids (TDS) continued to be far below safety limits (210–250 ppm), indicating that the water's mineral content was quite low. Electrical conductivity (EC) measurements fell within acceptable ranges(290-350 μS/cm), indicating suitable groundwater quality. Dissolved oxygen (DO) levels varied(3.7-9.3 mg/L), with some sites exhibiting satisfactory oxygenation while others raised concerns about potential water quality issues. Total hardness(TH)levels(330-900 mg/L)exceeded recommended limits(200-500 mg/L)at all sites, suggesting potential issues with water quality. However, chemical oxygen demand (COD) levels were minimal or zero(0-3 mg/L), indicating minimal organic pollution. Iron concentrations were generally low(0-0.05 mg/L).Total chlorine(TC)concentrations varied(0-0.03 mg/L)across sites.In summary, the study emphasizes how crucial it is to continuously monitor and regulate the quality of groundwater in order to guarantee the sustainability and safety of water resources in the Mirpur region
Integrated Employee Database System for South Sumatra’s Tourism and Culture Department
No full textThe management of employee data in many governmental organizations remains entrenched in
outdated, manual processes that are inefficient and prone to errors. This issue was notably present
at the South Sumatra Province Culture and Tourism Office, where the existing method of using
Microsoft Excel and Word for handling employee records was becoming increasingly untenable.
The manual system not only required excessive administrative effort but also exposed the
organization to significant risks of data loss and errors, which are detrimental to effective human
resource management. To address these challenges, a web-based information system was
developed using the PHP programming language. This system was designed to automate and
streamline the entire process of managing employee data, from entry to retrieval and reporting.
The system includes several key components: a secure login page, a profile page for quick access
to important data, dedicated pages for managing specific types of information such as employee
details and position data, and a comprehensive reporting page for generating actionable insights
from the data collected. The results of implementing this new system were transformative. It
significantly reduced the time and effort required to manage employee data, improved the accuracy
of the data stored, and enhanced the security measures protecting sensitive information. The
system's user-friendly interface and robust functionality were well-received by the staff,
facilitating smooth adoption and integration into daily operations. The new web-based information
system has successfully modernized the administrative functions of the South Sumatra Province
Culture and Tourism Office. It has established a more reliable, efficient, and secure framework for
managing employee data, setting a strong example for similar advancements in other governmental
departments. Future recommendations include ongoing updates to the system and continuous
training for users to ensure it continues to meet the evolving needs of the organizatio
The Role of Flexible Work Arrangements in Enhancing Employee Well- Being and Productivity: A Study on SDG 8 Implementation in Tech Companies
No full textEmployee well-being and productivity have come to the forefront of human resource
management (HRM) in the technology sector. This trend is coherent with the goals of
Sustainable Development Goal 8 (SDG 8), which promotes sustained, inclusive, and
sustainable economic growth, full and productive employment, and decent work for all
(Küfeoğlu, 2022). The high-pressure, fast-paced environments endemic to the tech industry
align with the intent to enhance workplace environments to promote mental health and overall
well-being in pursuit of SDG 8. According to Warmate et al. (2021), this mainly includes
allowing for flexible work arrangements (FWAs), such as remote work, flexible hours, and
compressed workweeks. The pandemic has forced the tech industry to adopt FWAs; in some
cases, this might even have been a positive development, particularly regarding employee
satisfaction and mental health, which are good for business. This study aims to address SDG 8
by examining the implementation of FWAs in tech companies as an avenue for improving the
well-being and performance of employees. The application of theories like Self-Determination
Theory (SDT) highlights how FWAs fulfil basic psychological needs such as autonomy,
competence, and relatedness, thereby boosting engagement and motivation among employees