INTERNATIONAL SCIENTIFIC JOURNAL OF ENGINEERING AND TECHNOLOGY (ISJET)

A Review of Pedestrian Information Retrieval Research

Yan Xie, Jian Qu — Mon, 30 Jun 2025 00:00:00 +0700

Pedestrian Information Search (PIS) has gained attention for its wide range of practical applications. The main objective of PIS is to find a matching object in a set of scene images or videos. Early work on PIS focused on image-based search. With the advent of deep neural networks, PIS can be freed from the limitations of the search source. Therefore, a systematic study of PIS is necessary. In this paper, we review the research results of PIS based on diﬀerent modalities in terms of the origin of the PIS task, the development history of PIS, and the methods of training and evaluation of PIS models. We selected the better-performing models for experiments. We summarize and comparatively evaluate the experimental results. Finally, we discuss some of the present problems of PIS and some meaningful future research directions.

Defect Reduction in Automotive Seat Manufacturing: A Lean Six Sigma Approach

Bundit Wongtong, Poom Jatunitanon, Bundit Inseemeesak, Yodnapha Ketmuang — Mon, 30 Jun 2025 00:00:00 +0700

This study investigates the issue of rear cushion wrinkling in a pickup truck seat production line using the Lean Six Sigma (LSS) methodology. By applying the DMAIC framework, we identified that excessive tension in the extruded listing ﬂeece caused deformation, particularly in curved seat sections. To resolve this problem, we redesigned the ﬂeece by incorporating rectangular slots (15 × 5 mm) spaced 80 mm apart. As a result, wrinkling defects were reduced by 60%, from 312 to 187 pieces, lowering the overall defect rate from 4.05% to 1.61% over six months. This exceeded our initial goal of reducing defects to less than 2.0%. Additionally, this improvement led to estimated cost savings of 852,500 THB, primarily due to a reduction in rework and material waste. Beyond cost benefits, the new design helped streamline the production process, cutting cycle time by 20% and improving customer satisfaction by a similar percentage. While these results demonstrate the eﬀectiveness of Lean Six Sigma in quality improvement, certain limitations remain. Factors such as operator variability and material inconsistencies were not fully controlled in this study. Future research could explore real-time defect detection systems or adaptive tension control mechanisms to enhance process stability.

Enhancing Thai Rice Query Assistance through a Knowledge-Driven Approach Using GraphRAG

Mon, 30 Jun 2025 00:00:00 +0700

Thai rice farmers face significant challenges accessing timely and accurate information for crucial decisions regarding variety selection, soil management, and adapting to climate change. While Retrieval-Augmented Generation (RAG) systems aim to provide information, traditional RAG often struggles with complex queries requiring interconnected knowledge and can yield generic or less relevant answers in specialized domains like agriculture due to its reliance on the semantic similarity of isolated text chunks. This paper introduces and evaluates GraphRAG, a knowledge graph-enhanced RAG approach, designed specifically to overcome these limitations and improve query assistance for Thai rice cultivation. The methodology involves constructing a knowledge graph from key Thai rice farming documents and integrating it with a large language model to provide context-aware responses, comparing its performance against a traditional RAG baseline. Results demonstrate GraphRAG’s superior effectiveness; user preference tests showed participants favored GraphRAG responses (52.9%) significantly more than traditional RAG (35.3%), particularly for complex queries requiring nuanced understanding. Quantitatively, GraphRAG showcased enhanced efficiency, reducing the average query response time by nearly 3 times (from 1.43 seconds for RAG to 0.41 seconds) and decreasing memory usage by over 50% (from 457.42 KB for RAG to 213.09 KB). This study concludes that GraphRAG offers a valuable approach for enhancing information retrieval accuracy, contextual understanding, and system efficiency in specialized, low-resource agricultural domains, highlighting its significance for providing better decision support to farmers.

Eﬀect of Packaging Materials on the Quality and Shelf Life of Fresh-Cut Vegetables

Chairat Burana, Phatcharee Kittisuban, Ruamporn Liamkaew, Gen ENDO — Mon, 30 Jun 2025 00:00:00 +0700

The fresh-cut products market has seen significant growth over the past decade, expanding from foodservice to retail shelves, convenience stores, and mobile fruit vans. Controlling temperature, atmosphere, relative humidity, and sanitation is crucial for maintaining the quality, safety, and shelf life of fresh-cut produce. The eﬀect of diﬀerent types of packaging on the quality and shelf life of fresh-cut lettuce, onion, and mixed cabbage with carrot was investigated. All fresh-cut produce was packaged in three types of packaging (A-Best®, Fresh & Fresh®, Active Pak®, and A-Best® at 5°C) and then stored at 4°C to simulate the refrigerated shelf in convenience stores. The result found that fresh-cut lettuce, onion, and mixed cabbage with carrot in A-Best at 5°C showed the highest overall acceptability both externally and internally, significantly. On the other hand, A-Best® and A-Best® at 5°C eﬀectively retarded the rate of browning symptom of the cut surfaces. The results suggest that combined treatments showed better results than those in the single treatment and have commercial potential in improving the shelf life and maintaining the quality of fresh-cut produce.

Intelligent Assessment of Athlete Physical Fitness: Addressing Data Imbalance

Mon, 30 Jun 2025 00:00:00 +0700

This study aims to mitigate the impact of imbalanced data through the use of the oversampling technique and to develop supervised learning models for assessing the physical fitness of youth athletes. The dataset comprises the physical fitness test results from 75 athletes aged 11 to 16 years. The dataset presents two major challenges: A limited sample size and a significant class imbalance, with certain fitness levels being underrepresented. This class imbalance can substantially degrade the performance of classification models, as it often leads to biased predictions favoring the majority class while failing to learn the characteristics of minority classes, those that may be most critical in practice. To address this issue, the Synthetic Minority Oversampling Technique (SMOTE) was employed to synthetically balance the class distribution. Five supervised learning algorithms were evaluated: Light Gradient Boosting Machine, Decision Tree, Random Forest, Neural Network, and Multinomial Logistic Regression. The Light Gradient Boosting Machine model yielded the highest accuracy at 87.76%, followed by Decision Tree, Random Forest, and Neural Network models, each with an accuracy of 79.59%. The Multinomial Logistic Regression model achieved the lowest accuracy at 75.51%. On average, the classification accuracy across all models improved to 81.41%, representing a 12.23% increase compared to using the original imbalanced dataset. The results demonstrate that applying oversampling techniques such as SMOTE can eﬀectively alleviate the eﬀects of class imbalance and enhance the predictive performance of machine learning models in the context of physical fitness assessment.

Predictive Analysis of Academic Achievement in Information Studies: A Comparative Study Using Educational Data Mining Techniques

Knitchepon Chotchantarakun — Mon, 30 Jun 2025 00:00:00 +0700

Predicting students’ academic achievement in the initial stages is beneficial for designing eﬀective training programs to enhance success rates. Extracting knowledge from student data is a fundamental aspect of Educational Data Mining (EDM). This study aims to analyze the predictive factors inﬂuencing the outcomes of graduates from the Information Studies program. The results not only contribute to improving student performance but also aid in constructing a better curriculum. A dataset is utilized within five classification models to categorize students into four target classes. The datasets are grouped into three types: Demographic information, course grades, and early-stage GPA. This study addresses the issue of the imbalanced dataset by applying the Synthetic Minority Over-sampling Technique (SMOTE). The findings indicate that early-stage GPA (90.5%) is the most significant predictor, particularly when applying the Naive Bayes classifier on a balanced dataset. In contrast, demographic information (58.0%) and core course grades (87.5%) show lower predictive inﬂuence. The findings support learning strategies and enhancing curriculum design to improve final academic outcomes.