Computer Science

Researchers from The University of Tokyo have found that single cells in collective chemotaxis act like agents in distributed reinforcement learning, utilizing the environment as an “external memory” and exhibiting highly intelligent behavior.

SUTD researchers test a privacy-preserving approach that blends federated learning with cloud coordination for smart grids in future 6G environments.

SOSE researchers are looking at a possible vaccine for H. pylori, a widely common bacterium that causes stomach ulcers and raises the risk for stomach cancer.

Lingnan University in Hong Kong has entered the Times Higher Education (THE) World University Rankings 2026. In the latest published results, Lingnan ranked among the world’s top 301–350 out of 2,191 universities from 115 countries or territories, delivering a notable performance across several indicators. These include an impressive 47th place in International Outlook, recognising Lingnan’s achievement in fostering international diversity and collaborative research on a global scale.

With so many environmental issues arising, how can we even begin trying to solve them all? Perhaps, we can start by asking AI. This study highlights how AI can pinpoint viable action plans for these issues.

Traffic accidents and last-minute service requests often disrupt logistics efficiency. Dr Tong Hao, Postdoctoral Fellow of the Division of Artificial Intelligence of the School of Data Science at Lingnan University, has developed an Intelligent Vehicle Routing Optimisation System capable of dynamically simulating traffic conditions and task variations to show the most efficient routes for vehicle fleets when faced with sudden congestion or additional deliveries, so as to maximise operational efficiency. His dissertation Advancing optimization and evaluation for dynamic capacitated arc routing problems is the only research project worldwide to receive the 2026 IEEE Computational Intelligence Society Outstanding PhD Dissertation Award.

The University of Osaka researchers developed a reinforcement learning framework that lets dialogue systems efficiently learn new words while asking fewer questions. This approach enhances user experience and enables future systems to naturally acquire family-specific nicknames and expressions, becoming more familiar companions in everyday life.

Lightweight and high-performance framework for remote sensing object detection

Lingnan University has achieved another remarkable milestone in the 2025 edition of the World’s Top 2% Scientists list, published by Stanford University. This year, a total of 43 Lingnan scholars were named in the list, representing an increase of about 50 per cent compared with last year’s 29. Among them, 19 scholars were ranked among the global top 100 in their respective disciplines, underscoring the international recognition of Lingnan’s research excellence and its rising global impact.

Leading scholar in artificial intelligence (AI) Prof Sam Kwong Tak-wu, Associate Vice-President (Strategic Research), Dean of the School of Graduate Studies and J.K. Lee Chair Professor of Computational Intelligence at Lingnan University, has been honoured with the 2026 IEEE Transactions on Evolutionary Computation Outstanding Paper Award for his co-authored paper Learning-Aided Evolution for Optimization. It is the sole paper to receive this award this year, underscoring its importance in the fields of evolutionary computation and AI.

A joint team has uncovered how soft, deformable particles, like cells, behave in microfluidic channels. Using precisely fabricated hydrogel particles and simulations on the supercomputer "Fugaku," they demonstrated that particle softness dramatically alters their focusing patterns, deviating significantly from rigid particle behavior. These findings reveal distinct "phase transitions" in focusing, shifting from mid-edge to eight-point, diagonal-edge, and finally center focusing as deformability increases. This breakthrough, explained by a new theoretical model incorporating inertia and deformability, offers crucial insights for designing next-generation microfluidic devices for highly efficient cell sorting and other biomedical applications like early cancer detection. The ability to control particle focusing based on deformability opens exciting possibilities for advanced particle manipulation and separation technologies.

Increases in wildfires, linked to projected future climate change, are expected to cause a notable increase in premature deaths, according to two new modelling studies.

Lingnan University held its Research and Knowledge Transfer Excellence Awards Presentation Ceremony 2025 yesterday, 17 September, honouring 55 faculty members for their outstanding contributions to academic research and knowledge transfer, advancing community well-being and societal progress during the past academic year. The awards included the Outstanding Researcher Award, the Early Career Researcher Award, and the Research & Knowledge Transfer Fund Award, recognising faculty members for their innovative scholarship and impactful initiatives that elevate Lingnan’s global research presence.

Lingnan University President and Wai Kee Kau Chair Professor of Data Science, Prof S. Joe Qin, welcomed the 2025 Policy Address delivered by Chief Executive, Mr John Lee, highlighting a series of education-related policy initiatives as follows:

Researchers at The University of Osaka and Kanazawa University have developed a novel method for analyzing cancer metabolism, revealing new insights into cancer's inefficient energy process. This breakthrough, published in Metabolic Engineering, combines biological experiments with advanced information science techniques to uncover the role of cancer-specific inefficient metabolism.

The University of Osaka D3 Center will begin trial operations of the “Osaka University Compute and sTOrage Platform Urging open Science” (OCTOPUS), a computational and data platform promoting open science built by NEC Corporation (NEC; TSE: 6701), starting this September, with full-scale operations commencing in December.

Perimetry (visual field testing) quantifies a patient’s retinal sensitivity to light and clarifies a deviation from normal retinal sensitivity. Visual field tests generally require high patient concentration, which can be exhausted. We constructed a framework for deep reinforcement learning to train ViFT (Visual Field Transformer), which controls all processes of visual field testing. ViFT achieves the same or higher accuracy than the state-of-the-art strategies, with less than half the test time of the state-of-the-art strategies.

Around 354 million people in North America and Europe are likely to have been affected by fine particulate matter pollution from the 2023 Canadian wildfire season, a new study suggests.

- A research team led by Professor Kyung-In Jang at DGIST has developed a flexible implantable device for brain drug delivery, which applies a pump structure inspired by human gastrointestinal motility. - Wireless control enables precise drug delivery to deep brain regions, thus opening new possibilities for treating intractable brain diseases.

Taiwan Innotech Expo (TIE) 2025 will return from October 16–18 at Taipei World Trade Center, Hall 1, uniting government, academia, and industry under the theme “Cross-Domain AI Innovation: Driving the Future with Intelligence.” The event will feature 450 exhibitors presenting nearly 1,100 patented technologies, highlighting Taiwan’s role as a key driver of global innovation.

Researchers at the University of Osaka have developed NeuraLeaf, a revolutionary CG model using deep learning to represent diverse plant species and their leaf deformations. This single model overcomes the limitations of traditional manual modeling by disentangling species-specific shapes from dynamic 3D deformations like wilting or curling. NeuraLeaf allows precise tracking of leaf changes, enhancing growth prediction, disease detection, and agricultural management. Presented at ICCV 2025, this technology promises to advance plant science and contribute to "PlantTwin," a project creating digital twins of plants.

This news article reports on a study revealing the importance of the proboscis monkey's large nose in vocal communication. Researchers used CT scans and computer simulations to demonstrate that the nose's shape modifies the resonant frequencies of calls, creating unique vocal signatures for individual monkeys. This discovery highlights the role of the nose in enhancing vocal identity and provides insights into the evolution of communication, not just in proboscis monkeys but potentially in other species as well. The collaboration between scientists and the Yokohama Zoo Zoorasia opens exciting new avenues for understanding the link between physical traits and social behaviors in the animal kingdom.

Combining multiple sensory inputs has been a big hurdle for developing physical AI robots. Recent research at Tohoku University solves this problem using a brand new approach.

Lingnan University congratulates the “Shenzhen–Hong Kong–Guangzhou” cluster on achieving the top position in the 2025 Global Innovation Index (GII) Ranking of World’s Top 100 Innovation Clusters. Released by the World Intellectual Property Organization (WIPO), this ranking recognises the Greater Bay Area’s (GBA) outstanding innovation capacity and its vital strategic role in the global innovation ecosystem.

A new study by an SUTD researcher and his collaborators introduces a pooled mining attack that overturns a long-standing assumption about Bitcoin’s security economics.

The School of Graduate Studies (GS) of Lingnan University in Hong Kong hosted the “Empower Your Future” Lingnan Postgraduate Orientation 2025 today, 28 August, welcoming about 3,600 new postgraduate students for the 2025/26 academic year, the highest number ever recorded. The cohort includes students from the Hong Kong SAR, Macau, Mainland China, Canada, France, Portugal, Switzerland, the US, Vanuatu, Ghana, and Cameroon enrolled in taught master’s, doctoral, and MPhil/PhD programmes.

Researchers at Tohoku University have unified experimental and computational data into one valuable resource: a map that instantly guides scientists to suitable materials out of countless candidates.

Lingnan University held its 2025/26 academic year New Student Orientation (NSO) and Official Welcome on campus today, 25 August, welcoming over 1,400 new undergraduates from the Hong Kong SAR, the Mainland, and overseas, including students from Germany, Kazakhstan, Nepal, Russia, South Korea, and Vietnam into the Lingnan family.

On 19 August, World Humanitarian Day, Lingnan University announced the signing of an Agreement of Cooperation with the United Nations University (UNU) to establish the UNU Hub on Humanitarian Innovation and Technology at Lingnan University (UNU Hub at Lingnan University), making it the first higher education institution in Asia to set up a UNU Hub. The Hub is scheduled to open officially in October 2025 and will have an annual operating budget of over HK$8 million. Its Humanitarian Innovation and Technology Laboratory (HIT Lab) will spearhead research and design initiatives, education and training programmes, and international collaborations, and improve living conditions in developing countries and regions of the Global South by leveraging innovative and cutting-edge technologies. Scholarships will be awarded to encourage local youth to fulfil their ambitions through knowledge and innovation.

Sensitive region detection for improved non-invasive cattle monitoring