Computer Science

To automatically generate data for training deep convolutional neural network models to segment building facades, researchers from Osaka University used a three-dimensional model and game engine to generate digital city twin synthetic training data. They found that a model trained on these data mixed with some real data was competitive with a model trained on real data alone, revealing the potential of digital twin data to improve accuracy and replace costly manually annotated real data.

When the evolution of towns and of roads are modeled together, the natural landscape alone is enough to predict the actual arrangement of real towns.

A pioneering and bold attempt to apply artificial intelligence (AI) technologies in the performing arts was carried out by Hong Kong Baptist University (HKBU) in a public concert held on 14 July in Hong Kong. The innovative performance was the first human-AI collaboration of its kind in the world, and it showcased how AI can be a creative force that can perform music, create cross-media art, and dance.

The Hong Kong Baptist University Symphony Orchestra (HKBU Symphony Orchestra) will present an innovative performance showcasing human creativity alongside artificial intelligence at its Annual Gala Concert, to be held on14 July (Thursday) at 8:00pm in the Hong Kong City Hall.

Atomic force microscopy (AFM) allows to visualize the dynamics of single biomolecules during their functional activity. All observations are, however, restricted to regions accessible by a fairly big probing tip during scanning. Hence, AFM images only the biomolecular surface with limited spatial resolution, missing important information required for a detailed understanding of the observed phenomena.

A research team has studied how humans react to the introduction of AI decision making. Specifically, they explored the question, “is society ready for AI ethical decision making?” by studying human interaction with autonomous cars.

The new multi-tasking weather sensor simultaneously measures rain volume and wind speed. Incorporating machine learning to analyze the output data, a single sensor can provide localized weather data in a timely manner, improving disaster preparedness.

199 Gold, 188 Silver and 56 Bronze were awarded

Neurone-like junctions made of mixed oxide-based materials could reduce the massive energy consumption of artificial intelligence operations.

The SupercomputingAsia Conference recognised key HPC pioneer-leaders with its annual SCA22 Awards. The winning teams from Asia, Australia and the US were also announced at the awards ceremony for the international Data Mover Challenge 2021 (DMC21) competition.

Computer game uses a camera and AI-based movement recognition to teach sign language to a broader public.

Just like explorers need maps, scientists require guides to better understand and advance new technology. A neuromorphic device, which can mimic the neural cells in our brain, has lacked such a guideline and created headaches for scientists trying to understand their operational mechanisms. That is until now after a research group created a map that provides rational design guidelines for neuromorphic devices, paving the way for advancements in brain-inspired computers.

Scientists predict that continued global warming under current trends could lead to an elevation of the sea level by as much as five meters by the year 3000 CE.

Despite significant advancements in computing technologies, they pale in comparison to the energy efficiency of human brains. Now, researchers from Tohoku University and the University of Gothenburg have developed a new spintronic technology for brain-inspired computing.

In a continuing effort to improve upon previous work, a research team at the Graduate School of Science, Osaka City University, have applied their recently developed Bayesian phase difference estimation quantum algorithm to perform full configuration interaction (full-CI) calculations of atoms and molecules without simulating the time evolution of the wave function conditional on an ancillary qubit. Superior to conventional methods in terms of parallel execution of quantum gates during quantum computing, this new algorithm is expected to be much easier to implement in actual quantum computers.

Prof. Jae Youn Hwang, DGIST, developed a domain-adaptive deep learning technology for object segmentation with state-of-the-art performance

A new agreement between Japan’s Research Organization for Information Science and Technology (RIST) and National Supercomputing Centre (NSCC) Singapore opens the path for Singapore researchers to regularly access and request for supercomputing resources from the world’s most powerful supercomputer, Japan’s Fugaku system. In the first such arrangement of its kind outside of Japan, Singapore researchers can now apply for available resources through an annual Call for Projects to Fugaku, which will be coordinated by NSCC and RIST.

Scientists develop new algorithm for rapid, computerized diagnosis of COVID-19, overcoming the limitations of reverse transcription polymerase chain reaction

Scientists propose an advanced network architecture that promises to overcome the drawbacks of traditional networks.

As part of its aim to improve business and government agency efficiencies through the use of Information and Communications Technology (ICT), the Advanced Science and Technology Institute (DOST-ASTI), a research and development institute of the Department of Science and Technology, has signed a 3-year partnership with Bizooku Philippines on 27 August 2021.

An algorithm that detects events occurring across different time and geographical scales on social media could enable a more timely response to emerging events.

Researchers have developed a general quantum algorithm that can directly calculate the energy difference of an atom and molecule using a quantum computer. By avoiding the need to calculate the total molecular energies, the general algorithm is expected to be applied not only to quantum chemical calculations but also to various physical and mathematical problems, which are intractable with nowadays classical computers.

Prof. Jae Youn Hwang's team, DGIST, develops deep learning technology to segment buildings from aerial images

Scientists create a new process control model to protect systems from assailants and natural fluctuations

Development of a deep learning technique that reduces the difference between image domains by separating and converting domain information of image capturing environments

Combining two magnetic effects can manipulate the magnetic behaviours of a device, offering a novel opportunity for the next generation IT technologies.

Researchers from The University of Tokyo Institute of Industrial Science develop an experimentally supported mathematical model that defines the velocity jump mechanism in crack propagation

Scientists design an innovative file system that overcomes current performance bottlenecks when using solid state drives

M.S. Student Sejin Kim & Jungwoo Kim, won the “Outstanding Paper Award” at the largest computer society conference

Recognized for its Research Outcome Related to PIM-based AI Processor Using TSMC’s Multi-bit RRAM