39506 research outputs found
Sort by
Review of Deepfake Detection Techniques and Challenges
No full textThe proliferation of deepfake technology, powered by advanced generative models such as generative adversarial networks (GANs), presents challenges in digital media authenticity, public trust, and cybersecurity. We reviewed recent advancements in deepfake detection across multiple modalities, including image, video, and audio. Benchmark datasets, such as FaceForensics++, the deepfake detection challenge (DFDC), and Celeb-deepfake (Celeb-DF), have been used to develop diverse detection models. These models encompass approaches based on EfficientNet-driven transfer learning, convolutional neural network–long shortterm memory (CNN-LSTM) hybrids for temporal feature extraction, graph-based neural architectures, and ensemble methods that integrate deep learning with handcrafted features. Although certain models report detection accuracies as high as 99.99% on specific datasets, many exhibit limited generalizability across different benchmarks, particularly when confronted with compression artifacts. Additionally, real-time deployment remains constrained by substantial computational
demands. Emerging threats, including adversarial perturbations and diffusion-based synthetic media, necessitate the development of more resilient detection strategies. Proactive countermeasures such as blockchain-based timestamping, digital watermarking, and cryptographic hashing have been adopted to enhance media integrity. The results of the review underscore the need for lightweight, interpretable, and multimodal detection frameworks to generalize the models’ applicability across diverse domains, thereby supporting reliable and scalable media verification in increasingly complex digital environments
Varanus salvator (Common Water Monitor Lizard). Diet
No full textVaranus salvator is a large varanid, with a distribution from the Indian Subcontinent to mainland and the western portion of insular Southeast Asia. It is a dietary generalist and an opportunistic predator, known to feed on a diverse range of prey including crabs and other invertebrates, fish, frogs, turtles, snakes, birds, small to mid-sized mammals, the eggs of reptiles and birds, as well as scavenge on animal carcasses and human- generated garbage. Here, we report on a novel frog in the diet of V. salvator from East Malaysia (North- western Borneo)
Fostering creativity and self-efficacy through collaborative learning using generative Artificial Intelligence (AI) in product design visualization process
No full textGenerative models in Artificial Intelligence (AI) are increasingly employed across diverse fields, including product design, for tasks like shape recognition and design creation. This trend underscores generative models' ability to bridge offline and online environments in creative endeavors. The article investigates the potential of integrating generative image AI into visualization process among product design students. Using image-based research analysis and semi-structured interviews, this study involved 50 product design students as respondents. The findings highlight that integrating generative AI tools, particularly the ChatGPT 4.0, significantly im�proves students' creativity and self-efficacy through collaborative learning, and streamlines the design process. The findings also close the gap between creative concepts and practical applications, and offers a robust framework for evaluating AI-generated content. The contribution of the study underscores the transformative potential of generative AI tools in product design education, showcasing the effectiveness in fostering creativity, efficiency, and design quality through collaborative learning
Re-reading Room (Kuching Chinese Voices on Heritage) Exhibition 2025
No full texthis "Re-Reading Room" features two interconnected parts.
The first part is the cave-shaped installation right in front of you. In creating this massive work, architect and artist Tay Tze Yong combined rigorous architectural engineering with his wild imagination of space and urban life. Paying tribute to the ancient tradition of co-creating art in caves, Tay invited people from different walks of life in Kuching to contribute their creativity to this
work. This is why the installation is named "Echoes of Creation." The second part is the reading space we have curated for you. In this room, we invite you to re-read the Chinese communities of Kuching and their participation in heritage activities. Do you think of temples, festival parades, and Carpenter Street when you hear "Kuching Chinese"? Do you visualise traditional crafts,
archaeological sites, and artefact conservation when you hear the word "heritage"? While you are absolutely right, this Re-Reading Room reveals much more than that.
How are these two parts connected? During the installation's
display at Hoan Gallery from June 8 to 19 August 2025, local
Chinese associations, schools, and individuals were invited to participate in its co-creation. Some of their contributions reflect their perceptions of Chinese identity and what heritage means to them
UNIMAS researcher’s cancer study gains international recognition
No full textThis UNIMAS news article reports the international recognition received by Dr. Tan Sang Nee for her research in cancer immunology. Her study, published in the Journal of Experimental Medicine (JEM) and selected for both the JEM Cancer Collection and the JEM Clinical Immunology Collection, uncovers how Th1 cells convert into regulatory T cells in the tumor microenvironment via CD39. The article also highlights her media coverage by OncoDaily, recognition by Sigma Xi, and her scientific contributions from postgraduate training at UNIMAS through her postdoctoral work at Tsinghua University
Corporate Governance, Risk Management, and Financial Performance: The Mediating Role of Risk Governance in Indonesia's Islamic Banks
No full textIslamic banks face increasing pressure to strengthen governance and risk oversight, yet the mediating role of risk governance infinancial performance remains underexplored, particularly in Indonesia. This study aims to examine the structural relationships among corporate governance, risk management, risk governance, and the financial performance of Indonesia’s Islamic banks. A quantitative research, drawing on secondary data from annual reports of eight Islamic banks between 2016 and 2023with64 observations, was employed. Corporate governance was measured using a composite governance index, risk governance through committee structures, risk management via credit, operational, liquidity, and Sharia-compliance controls, while financial performance was proxied by ROA and ROE. Data were analyzed using path analysis with PLS-SEM. The findings reveal that corporate governance significantly strengthens risk governance but does not directly improve financial performance. Conversely, risk management exerts a significant positive influence on financial performance, though it does not shape risk governance. Risk governanceitself neither impacts financial performance nor mediates the relationships between governance, risk management, and performance. These results suggest that risk governance in Indonesian Islamic banks remains compliance-oriented rather than performance-enhancing. The study highlights the need for regulatory reforms and strategic integration of risk governance to support sustainable growt
Enhancing wastewater treatment efficiency by using ammonium-based aeration control technology for energy reduction and effluent quality improvement
No full textMotivated by the increasing operational costs and stringent effluent requirements faced by wastewater treatment plants (WWTP) operating activated sludge systems, this study
explores the potential of ammonium-based aeration control (ABAC) technology to enhance the current dissolved oxygen (DO) control strategies. The primary objectives are to reduce energy consumption and improve effluent quality to meet regulatory standards. In this study, ABAC and
nitrate (SNO2) control were proposed and analysed using the benchmark simulation model no. 1 (BSM1) simulation model. Both ABAC and nitrate control strategies were designed using neural network (NN). Comparative analysis against standard BSM1 DO proportional integral (PI) and SNO2 PI, as well as NN ABAC combined with SNO2 PI, reveals the superior effectiveness of the proposed control configuration. Specifically, it achieves the lowest violations of total nitrogen, ammonia, and suspended solids limits, highlighting its potential to improve the effluent quality of
the WWTP. The proposed strategy shows the most significant improvements across energy consumption, effluent quality, operational costs, and regulatory compliance
WORKING CAPITAL DYNAMICS AND FIRM PERFORMANCE: EVIDENCE FROM MALAYSIAN TECHNOLOGY FIRMS
No full textThis study examines how working capital management influences the financial performance of technology firms listed in the FTSE Bursa Malaysia Top 100 Index, addressing gaps in sector-specific evidence in Malaysia. Based on recent empirical findings linking WCM practices to profitability, we investigate the effects of the Cash Conversion Cycle, Average Payment Period, Current Ratio, and Leverage in relation to Return on Assets and Return on Equity. Using panel data from 12 firms over 2019–2023, and panel data regression, the Fixed Effects Model (FEM) shows that only LEV significantly and positively affects both ROE and ROA, supporting the Trade-Off Theory. To address endogeneity, the Generalised Method of Moments (GMM) confirms LEV’s significance and further indicates that CCC and CR also influence ROE when dynamic factors are considered. The findings suggest that financial leverage is the primary driver of performance, while efficient WCM enhances shareholder returns under advanced modelling. These findings highlight the critical role of leverage management as part of working capital strategy for Malaysian technology firms
A Study on the Mechanism of the Impact of Multidimensional Leadership Styles on Employee Innovation Behavior Based on Structural Equation Modeling Analysis
No full textTraditional people-oriented management approaches emphasize placing employees at the core while neglecting the impact of employees' negative emotions. Given this, how to effectively manage subordinates' negative emotions and stimulate employees' innovative behavior has become an urgent issue for both academia and managers to address. This study proposes five research hypotheses based on the theoretical framework of leadership style and employee innovative behavior, utilizing structural equation modeling. Through empirical analysis, it measures the multidimensional leadership style and employee innovative behavior variables and tests each hypothesis individually. it was found that transformational leadership, transactional leadership, laissez-faire leadership, servant leadership and empowerment leadership all had a significant impact on employees' innovative behaviors. Among them, laissez-faire leadership showed a significant negative impact on employee innovation behavior (p=-0.584, p<0.001), while the other four leadership styles showed a significant positive impact
The impact of AI applications, information sharing, and supply chain resilience on agricultural supply chain performance
No full textIn the digital era, improving supply chain performance (SCP) has become increasingly important for agricultural enterprises confronting global competition and market volatility. These organizations need to manage networks comprising multiple stakeholders, diverse information flows, and complex logistics processes, while addressing unique challenges in the agricultural sector. This study examines the interrelationships among four latent variables using partial least squares structural equation modeling (PLS-SEM): artificial intelligence (AI) applications, information sharing (IS), and supply chain resilience (SCR) as independent variables, with SCP as the dependent variable. Six hypotheses were tested using data collected from 151 agricultural enterprises in China through a structured questionnaire. This study contributes to the literature on supply chain management by examining how the triad of AI applications, IS, and SCR influences agricultural SCP. It provides insights into potential approaches for enterprises to implement AI-driven solutions in addressing sector specific challenges like perishability management and multi-stakeholder coordination in volatile markets