MMU Press (Multimedia University)
Not a member yet
714 research outputs found
Sort by
Impact of Data Quality Types on Computational Time in Data Source Selection Using Ant Colony Optimization
No full textData quality varies dramatically from source to source, even within the same domain. Given these challenges, data source selection has emerged as a crucial step in information integration. It demands efficient and scalable approaches that can handle massive data volumes while ensuring the quality of results. Adapting the ACO algorithm to solve the data sources selection problems may lead to inconsistent computational time if the data sources provided are vary in quality. These challenges bring the issues of time consuming in selecting the required data sources. However, how much the computational time needed in solving the data sources selection is depending on the type of data quality. Hence, in this article, the impact of quality type of data towards computational time is examined in solving the data sources selection problems. For the methodology used, there are five steps need to be followed which are first collect data set, second import the data sources to the data sources selection model, third implement the ACO algorithm, fourth obtain the computational time and lastly compare the results. The experiment shows that low-quality data set achieve higher computational time compared to the high-quality data set which achieve the minimum computational time by 3.38 % faster. The results obtained in this experiment shown that the quality type of data has given an impact to the computational time of ACO algorithm. The results also clearly show the contribution of high-quality data set in minimizing computational time in the selection process. The validation on quality type of data with computational time is to clarify the importance of selecting a good quality data to save the computational time
Pavement Distress Analysis in Malaysia: A Novel DeepSeg-CrackNet Model for Crack Detection and Characterization Using Real-World Data
No full textPavement distress analysis plays a big role in keeping roads in good shape, especially in busy spots like Selangor and Kuala Lumpur, where heavy traffic and tropical weather make them wear out fast. This work introduces DeepSeg-CrackNet, a fresh hybrid deep learning model that uses Deep Gradient ResNet to spot cracks and a Residual block with a Modified Attention Mechanism to sort them into types, making it simpler to detect and label pavement damage. The model was trained on real data collected from Malaysian roads, with the CRACK500 dataset added in to cover more situations, and captured using a GoPro Hero 8 mounted on a vehicle, with GPS mapping keeping everything clear and easy to trace. DeepSeg-CrackNet performs really well—it hits a Mean IoU of 0.8388889 for segmentation and scores 85% accuracy in classifying cracks like alligator, longitudinal, and transverse, with precision ranging from 0.84 to 0.89, and recall between 0.80 and 0.96. It also measures cracks in meters or square meters, which helps in planning repairs smartly, like replacing big alligator cracks or sealing smaller longitudinal ones to save resources. Compared to models like CrackNet, DeepSeg-CrackNet stands out, especially for alligator cracks, with a precision of 0.84 and recall of 0.96, beating CrackNet’s 0.778 and 0.772. In the end, DeepSeg-CrackNet makes it easier to manage Malaysia’s roads in a data-driven way, improving safety and ensuring longer-lasting infrastructure through smarter, proactive repair approaches that enhance city travel
Design and Implementation of an Arduino-Based Ultrasonic Device for Humane Dog Repellent
No full textDogs can pose a nuisance and danger to people in residential areas through barking and territorial behaviour, causing discomfort and safety concerns. Dogs and humans both possess hearing capabilities, but dogs can detect ultrasonic frequencies that humans cannot perceive. This enhanced auditory sensitivity makes dogs responsive to high-frequency acoustic stimuli. In this paper, preliminary field observations of an Arduino-based ultrasonic dog deterrent device are presented to explore frequency response patterns in free-roaming dog populations. A frequency-based approach represents a potentially more environmentally safe alternative compared to traditional chemical repellents. This research presents observational data from field testing of a portable prototype that incorporates an Arduino Uno microcontroller, an ultrasonic transducer, and an amplifier to generate adjustable high-frequency sound waves. The microcontroller enables frequency control across the 38 to 42 kHz range to emit an ultrasonic sound that dogs respond to without physical harm. The device is portable, offers frequency adjustability, and is capable of field deployment. Based on the observations from forty encounters with stray dogs, the response rates increased across the frequency range. Across the frequency range, 42 kHz showed the highest observed response. These findings suggest that ultrasonic deterrent applications show promise. Further research is needed to confirm the effectiveness and optimal deployment parameters.
Manuscript received: 9 Jun 2025 | Revised: 14 Jul 2025 | Accepted: 21 Jul 2025 | Published: 30 Nov 202
A Reproducible Benchmark of AdamW-Augmented Lightweight Models for Trash Classification
No full textGlobal waste generation is projected to reach 3.40 billion tonnes by 2050, creating urgent demands for automated waste classification systems that can overcome the limitations of manual sorting methods. Current deep-learning research on waste classification lacks standardised evaluation protocols, preventing meaningful architectural comparisons and hindering the progress of reproducible research. This paper establishes a reproducible benchmark framework for lightweight neural network models designed explicitly for trash classification research applications. Lightweight models are designed for optmised architecture and computation cost while maintain accuracy. Four representative lightweight models, including MobileNet V3 Large, Vision Transformer (ViT) Small, EfficientFormer, and ShuffleNet V2, were systematically evaluated on the TrashNet dataset using identical training protocols. All models employed AdamW optimisation with a learning rate of 1 × 10-4, weight decay of 1 × 10-4, and CosineAnnealingLR scheduling through 5-fold stratified cross-validation on RTX 2080 Ti hardware. Experimental results demonstrate that ViT Small achieved the highest classification accuracy at 0.815 but required 21.67M parameters, while MobileNet V3 Large delivered superior computational efficiency with 0.768 accuracy and 0.72ms inference time using only 4.21M parameters. Statistical analysis revealed significant performance differences across models (p = 0.0002), with hardware-aware architectural optimisations proving more critical than raw parameter reduction for computational performance on data centre GPU hardware. The standardised evaluation framework and open-source implementation provide rigorous baselines for advancing automated waste classification research.
Manuscript received: 12 Jun 2025 | Revised: 7 Aug 2025 | Accepted: 11 Aug 2025 | Published: 30 Nov 202
Physical, Online, or Hybrid? A Study on the Preferred Mode of Learning of Multimedia University (MMU) Students
No full textHigher institutions in Malaysia, including Multimedia University (MMU), has adopted online and hybrid learning mode for its students during the Covid 19 pandemic and post pandemic. At present, this practice is still on going, apart from the physical learning mode norm. This sparks an interest in determining the preferred mode of learning of MMU students and is the basis for this study. A total of 363 respondents from different faculties across both Melaka and Cyberjaya Campus partake in this study. Several tests were conducted on the data collected, including Cronbach Alpha, Pearson Correlation, and multiple linear regression (MLR). Results indicated the survey instrument used was reliable across all variables. Weak relationships were found among all predictors to the three preferred learning modes. Albeit this, the MLR tests were conducted. In conclusion, upon comparing the results, it was determined that the preferred learning mode of MMU students is the online mode.
Manuscript received: 18 Jun 2025 | Revised: 30 Aug 2025 | Accepted: 5 Sep 2025 | Published: 30 Nov 202
An Edge Convolution Neural Network Model for Plant Health Classification Using Camera
No full textAs per the Food and Agricultural Organization (FAO), plant diseases infect approximately 1.3 billion tonnes of crops. Historically, farmers relied on visual inspection for disease detection and classification. In this study, a Convolutional Neural Network (CNN) with five convolutional layers was used to accurately recognize plant diseases. A deployable CNN model was developed for classifying plant diseases, integrated into a web application with a camera, forming a vision system integrated with CNN model. The CNN model was trained using a public dataset comprising 19,384 images of potatoes, peppers, and tomatoes, collected under controlled conditions. These plants were chosen due to their common occurrence in Malaysia. The evaluation metrics F1 score were used to assess the model’s performance. The accuracy and F1-score of the trained model were 97.2% and 97%, respectively.
Manuscript received: 26 Nov 2024 | Revised: 3 Jan 2025 | Accepted: 11 Jan 2025 | Published:: 31 Mar 202
Design of a Smart Surveillance Robot Using HUSKYLENS AI Vision Sensor
No full textThis study aims to enhance security by implementing a smart surveillance robot using existing facial recognition technology. The developed robot is equipped with the HUSKYLENS AI vision sensor camera and an ESP32 CAM module to provide a real-time video feed to a connected computer via Wi-Fi. The results demonstrate the successful integration of facial recognition technology into the surveillance robot's functionality. The robot exhibits an acceptable ability to identify and track intruders, underscoring its potential for enhancing security applications. The robot features a height of 12 cm, a width spanning 11.5 cm, a length measuring 19 cm, and a weight totalling 1.3 kg. Operating on a basic configuration of two main wheels, the robot forms a two-wheeled system with three degrees of freedom (DOF). The developed robot demonstrates an ability to identify and track intruders with a tested accuracy of 77.5%, precision of 80%, specificity of 79%, and sensitivity of 76.1%. The compact and low-profile design enables it to operate discreetly in diverse environments, making it particularly well-suited for scenarios where inconspicuous surveillance is needed.
Manuscript received:27 Feb 2025 | Revised: 29 Apr 2025 | Accepted: 7 May 2025 | Published: 30 Jul 202
Optimizing Reviewer Assignment with Recommender Systems: Models, Related Work, and Evaluation
No full textPeer reviewer assignment to academic articles is important in ensuring the quality and originality of academic publications. Traditional methods of selecting reviewers are generally plagued by inefficiency, reviewer burnout, and inconsistency between the subject of the manuscript and the reviewer area of expertise. In attempting to avoid such drawbacks, recommender systems have been explored as a means of solving the reviewer assignment problem. This article reviews the recommender system techniques in detail by reviewing their application in peer reviewer selection. Additionally, related works shall be examined for how different methods work, their strength and limitations, the dataset used by them, and evaluation metrics used in measuring system performance.
Manuscript received:11 Mar 2025 | Revised: 30 Apr 2025 | Accepted: 13 May 2025 | Published: 30 Jul 202
Women supporting women in entrepreneurship: Examining the role of women in empowering each other
No full textIn a world where women entrepreneurs continue to face systemic barriers, the support of other women entrepreneurs becomes not just a choice but a necessity for driving meaningful change and creating a more equitable and inclusive entrepreneurial landscape. Women entrepreneurs have experienced a significant 114% growth over the last twenty years, resulting in women owning one-third of businesses globally. Despite substantial progress, there is still a lack of knowledge regarding the successful development and support of women entrepreneurs. An important focus is analysing the impact of other women in supporting women's entrepreneurship. This study uses the social identity theory as the framework and conducts online in-person interviews with 17 women entrepreneurs from four locations in Malaysia: Melaka, Selangor, Johor, and Negeri Sembilan. The emphasis is on investigating the role of other women in the entrepreneurial process. The findings show five essential roles women entrepreneurs offer to other women to sustain their businesses in various settings. Compared to other roles, women serve as role models that are substantially more important. This study explores the broader consequences of supporting women in promoting women entrepreneuers
Potentialities of the environmental law and policy for e-waste recycling: a vista for sustainable development :
No full textE-waste production contains damaged, obsolete, non-functional, old, or expired goods, which are allocated into two sources: industrial and household production. Disposing e-waste will hurt public well-being. This study was conducted to examine the effects of environmental law and policy on the connections between environmental attitudes, subjective norms, perceived behavioral control, and e-waste recycling. The chosen study location was the southern region of Malaysia, and consisted of selected individual residents. The methodology used was quantitative, and this study used a questionnaire as the primary material; 258 respondents answered the survey. The data investigation method used was moderated regression analysis. This study concludes that environmental attitudes, subjective norms, and perceived behavioral control positively affect e-waste recycling behavior. Interaction investigation in regression shows that environmental laws and regulations can develop the control of environmental attitudes and subjective norms on e-waste recycling behavior. In contrast, environmental laws and regulations do not moderate the correlation between perceived behavioral control and e-waste recycling. Thus, the Malaysian government requires a strong legal and institutional structure for environmental protection and self-control practices to foster sustainable development