Communication & Information

Researchers developed an automated system to help programmers increase the efficiency of their deep learning algorithms by simultaneously leveraging two types of redundancy in complex data structures: sparsity and symmetry.

A study suggests that a computer-generated breathing coach could be as effective as sessions with a human trainer.

Researchers developed a new optical system that uses holograms to encode information, creating a level of encryption that traditional methods cannot penetrate.

A research team has introduced a new concept for designing ultra-thin absorbers that enables absorbing layers with a record-high bandwidth-to-thickness ratio, potentially several times greater than that of absorbers designed using conventional approaches.

Various methods are used to correct errors in quantum computers. Not all operations can be implemented equally well with different correction codes. Therefore, a research team has developed a method and implemented it experimentally for the first time, with which a quantum computer can switch back and forth between two correction codes and thus perform all computing operations protected against errors.

A scientific team looks at the progress and challenges in the research and development of brain implants. New achievements in the field of this technology are seen as a source of hope for many patients with neurological disorders and have been making headlines recently. As neural implants have an effect not only on a physical but also on a psychological level, researchers are calling for particular ethical and scientific care when conducting clinical trials.

While entangled photons hold incredible promise for quantum computing and communications, they have a major inherent disadvantage. After one use, they simply disappear. In a new study physicists propose a new strategy to maintain communications in a constantly changing, unpredictable quantum network. By rebuilding these disappearing connections, the researchers found the network eventually settles into a stable -- albeit different -- state.

Experts highlight the need for a clear framework when it comes to AI research, given the rapid adoption of artificial intelligence by children and adolescents using digital devices to access the internet and social media.

A brain-computer interface, surgically placed in a research participant with tetraplegia, paralysis in all four limbs, provided an unprecedented level of control over a virtual quadcopter -- just by thinking about moving his unresponsive fingers.

Scientists have developed a computing chip that can learn, correct errors, and process AI tasks.

Researchers applied the mathematical theory of synchronization to clarify how recurrent neural networks (RNNs) generate predictions, revealing a certain map, based on the generalized synchronization, that yields correct target values. They showed that conventional reservoir computing (RC), a type of RNN, can be viewed as a linear approximation, and introduced a 'generalized readout' incorporating further order approximations. Using a chaotic time-series forecasting task, they demonstrated that this approach dramatically enhances both prediction accuracy and robustness.

A new initiative is challenging the conversation around the direction of artificial intelligence (AI). It charges that the current trajectory is inherently biased against non-Western modes of thinking about intelligence -- especially those originating from Indigenous cultures. Abundant Intelligences is an international, multi-institutional and interdisciplinary program that seeks to rethink how we conceive of AI. The driving concept behind it is the incorporation of Indigenous knowledge systems to create an inclusive, robust concept of intelligence and intelligent action, and how that can be embedded into existing and future technologies.

AI applications like ChatGPT are based on artificial neural networks that, in many respects, imitate the nerve cells in our brains. They are trained with vast quantities of data on high-performance computers, gobbling up massive amounts of energy in the process. Spiking neurons, which are much less energy-intensive, could be one solution to this problem. In the past, however, the normal techniques used to train them only worked with significant limitations. A recent study has now presented a possible new answer to this dilemma, potentially paving the way for new AI methods that are much more energy-efficient.

Engineers have demonstrated a well-known quantum thought experiment in the real world. Their findings deliver a new and more robust way to perform quantum computations and they have important implications for error correction, one of the biggest obstacles standing between them and a working quantum computer.

Are humans or machines better at recognizing speech? A new study shows that in noisy conditions, current automatic speech recognition (ASR) systems achieve remarkable accuracy and sometimes even surpass human performance. However, the systems need to be trained on an incredible amount of data, while humans acquire comparable skills in less time.

Researchers have developed IRIS, a smart ring that allows users to point and click to control smart devices. The prototype Bluetooth ring contains a small camera which sends an image of the selected device to the user's phone. The user can control the device clicking a small button or -- for devices with gradient controls, such as a speaker's volume -- rotating the ring.

Imagine a future where your phone, computer or even a tiny wearable device can think and learn like the human brain -- processing information faster, smarter and using less energy. A breakthrough approach brings this vision closer to reality by electrically 'twisting' a single nanoscale ferroelectric domain wall.

Quantum computers require extreme cooling to perform reliable calculations. One of the challenges preventing quantum computers from entering society is the difficulty of freezing the qubits to temperatures close to absolute zero. Now, researchers have engineered a new type of refrigerator that can autonomously cool superconducting qubits to record low temperatures, paving the way for more reliable quantum computation.

AI tools are increasingly being used to track and monitor us both online and in-person, yet their effectiveness comes with big risks. Computer scientists have developed a new mathematical model which could help people better understand the risks posed by AI and assist regulators in protecting peoples' privacy.

AI-powered algorithm can analyze video recordings of clinical sleep tests and more accurately diagnose REM sleep behavior disorder.

Researchers have found a low-power, inexpensive way for large numbers of devices, such as machines in factories and equipment in labs, to share information by efficiently using signals at untapped high frequencies. The technology is an advanced version of a device that transmits data in a wireless system, commonly known as a tag.

Neural implants contain integrated circuits (ICs) -- commonly called chips -- built on silicon. These implants need to be small and flexible to mimic circumstances inside the human body. However, the environment within the body is corrosive, which raises concerns about the durability of implantable silicon ICs. A team of researchers address this challenge by studying the degradation mechanisms of silicon ICs in the body and by coating them with soft PDMS elastomers to form body-fluid barriers that offer long-term protection to implantable chips. These findings not only enhance the longevity of implantable ICs but also significantly broaden their applications in the biomedical field.

In order to uncover the relationship between structure and function, researchers used microfluidic devices to study neuronal networks.

Researchers have harnessed artificial intelligence to take a key step toward slashing the time and cost of designing new wireless chips and discovering new functionalities to meet expanding demands for better wireless speed and performance.

Artificial intelligence has the potential to improve the analysis of medical image data. For example, algorithms based on deep learning can determine the location and size of tumors. This is the result of AutoPET, an international competition in medical image analysis. The seven best autoPET teams report on how algorithms can detect tumor lesions in positron emission tomography (PET) and computed tomography (CT).

A research group has analyzed the digital ecosystem of 11- to 12-year-old children across the Basque Autonomous Community, and concluded that two out of three own a smartphone. They use smartphones mainly to talk to family and friends. The researchers also point out that, at that age, access to social media mainly focuses on watching videos and not on generating content.

Quantum teleportation could provide near-instant communication over long distances. But, inside Internet cables, photons needed for teleportation are lost within the millions of light particles required for classical communications. A new study quantified light scattering to find exact areas to place photons to keep them safe from other particles. The approach successfully worked in experiments carrying regular Internet traffic.

Artificial intelligence that is as intelligent as humans may become possible thanks to psychological learning models, combined with certain types of AI.

SMART, a new software package, can make studying signaling processes significantly easier. Results could accelerate research in fields across the life sciences, such as systems biology, pharmacology and biomedical engineering.

Researchers developed a laser-based artificial neuron that fully emulates the functions, dynamics and information processing of a biological graded neuron, which could lead to new breakthroughs in advanced computing. With a processing speed a billion times faster than nature, chip-based laser neuron could help advance AI tasks such as pattern recognition and sequence prediction.

Reinforcement Learning, an artificial intelligence approach, has the potential to guide physicians in designing sequential treatment strategies for better patient outcomes but requires significant improvements before it can be applied in clinical settings, finds a new study.

A study is the first-of-its-kind to recognize American Sign Language (ASL) alphabet gestures using computer vision. Researchers developed a custom dataset of 29,820 static images of ASL hand gestures. Each image was annotated with 21 key landmarks on the hand, providing detailed spatial information about its structure and position. Combining MediaPipe and YOLOv8, a deep learning method they trained, with fine-tuning hyperparameters for the best accuracy, represents a groundbreaking and innovative approach that hasn't been explored in previous research.

A new AI image tool could aid the development of algorithms to analyse wildlife images to help improve understanding of how species around the world are responding to climate change, a study suggests.

Researchers developed an AI debiasing technique that improves the fairness of a machine-learning model by boosting its performance for subgroups that are underrepresented in its training data, while maintaining its overall accuracy.

Researchers have developed a 3D-printed device that generates twisting light beams with orbital angular momentum (OAM), a form of rotational energy that can carry more data than regular beams.

Researchers have realized a new design for a superconducting quantum processor, aiming at a potential architecture for the large-scale, durable devices the quantum revolution demands. Unlike the typical quantum chip design that lays the information-processing qubits onto a 2-D grid, the team has designed a modular quantum processor comprising a reconfigurable router as a central hub. This enables any two qubits to connect and entangle, where in the older system, qubits can only talk to the qubits physically nearest to them.

When a quantum computer processes data, it must translate it into understandable quantum data. Algorithms that carry out this 'quantum compilation' typically optimize one target at a time. However, a team has created an algorithm capable of optimizing multiple targets at once, effectively enabling a quantum machine to multitask.

A groundbreaking piece of navigation technology that uses the ability to sense information through touch can help people with visual impairment perform a location task as well as sighted people, according to new research.

Physicists have devised an algorithm that provides a mathematical framework for how learning works in lattices called mechanical neural networks.

A research team has taken inspiration from the brains of insects and animals for more energy-efficient robotic navigation.

Since the first fiber optic cables rolled out in the 1970s, they've become a major part of everything from medical devices to high-speed internet and cable TV. But as it turns out, one group of marine mollusks was way ahead of us. A new study reveals that clams called heart cockles have unique structures in their shells that act like fiber optic cables to convey specific wavelengths of light into the bivalves' tissues.

Researchers developed a fully integrated photonic processor that can perform all the key computations of a deep neural network on a photonic chip, using light. This advance could improve the speed and energy-efficiency of running intensive deep learning models for demanding applications like lidar, astronomical research, and navigation.

Humans get a real buzz from the virtual world of gaming and augmented reality but now scientists have trialled the use of these new-age technologies on small animals, to test the reactions of tiny hoverflies and even crabs. In a bid to comprehend the aerodynamic powers of flying insects and other little-understood animal behaviors, the study is gaining new perspectives on how invertebrates respond to, interact with and navigate virtual 'worlds' created by advanced entertainment technology.

A new article examines the convergence of physics, chemistry, and AI, highlighted by recent Nobel Prizes. It traces the historical development of neural networks, emphasizing the role of interdisciplinary research in advancing AI. The authors advocate for nurturing AI-enabled polymaths to bridge the gap between theoretical advancements and practical applications, driving progress toward artificial general intelligence.

Deforestation has remained a significant issue globally, with primary forests contributing to 16 per cent of the total tree cover loss in the last two decades, driven by climate change and intensive human activity. This threatens natural resources, biodiversity, and people's quality of life. To protect forests, scientists have developed Forest 4.0, an intelligent forest data processing model integrating blockchain, Internet of Things (IoT), and Artificial Intelligence (AI) technologies. The system enables real-time monitoring of forest conditions, sustainable resource accounting, and a more transparent forest governance model.

The genome has space for only a small fraction of the information needed to control complex behaviors. So then how, for example, does a newborn sea turtle instinctually know to follow the moonlight? Neuroscientists have devised a potential explanation for this age-old paradox. Their ideas should lead to faster, more evolved forms of artificial intelligence.

To understand the practical challenges of indoor augmented reality applications on smartphones, researchers conducted 113 hours of extensive experiments and case studies over 316 patterns to determine the factors that degrade localization accuracy in real-world indoor environments. Landmarks for vision systems, LiDAR, and the IMU were evaluated. To solve the identified problems, the researchers suggest radio-frequency-based localization as a potential solution for practical augmented reality applications.

Researchers study the importance of enunciation when using speech-to-text software in medical situations.

Researchers have found a new way to improve a key element of thermophotovoltaic systems, which convert heat into electricity via light. Engineers designed a thermal emitter that can deliver high efficiencies within practical design parameters.

Archaeology studies the human past through the excavation of things people made and used thousands of years ago -- from architecture to objects like pottery bowls and animal bones from meals. Although many excavation projects create digital 3D models of what they uncover, archaeologists need new ways to meaningfully use those data. Some projects share 3D models with the public as tourism and teaching tools -- one may have recently seen 3D displays at museums. However, archaeologists themselves have not yet taken full advantage of these models in their own fieldwork and research. That is about to change.

Depression is one of the most common mental illnesses. As many as 280 million people worldwide are affected by this disease, which is why researchers have developed an artificial intelligence (AI) model that helps to identify depression based on both speech and brain neural activity. This multimodal approach, combining two different data sources, allows a more accurate and objective analysis of a person's emotional state, opening the door to a new phase of depression diagnosis.

In order to use remote locations to record and assess the behavior of wildlife and environmental conditions, the GAIA Initiative developed an artificial intelligence (AI) algorithm that reliably and automatically classifies behaviors of white-backed vultures using animal tag data. As scavengers, vultures always look for the next carcass. With the help of tagged animals and a second AI algorithm, the scientists can now automatically locate carcasses across vast landscapes.

When making diagnostic decisions, radiologists and other physicians may rely too much on artificial intelligence (AI) when it points out a specific area of interest in an X-ray, according to a new study.

Researchers have used machine learning and supercomputer simulations to investigate how tiny gold nanoparticles bind to blood proteins. The studies discovered that favorable nanoparticle-protein interactions can be predicted from machine learning models that are trained from atom-scale molecular dynamics simulations. The new methodology opens ways to simulate efficacy of gold nanoparticles as targeted drug delivery systems in precision nanomedicine.

A study has identified strategies for using large language models (LLMs), a type of artificial intelligence (AI), in health systems while maintaining cost efficiency and performance. The findings provide insights into how health systems can leverage advanced AI tools to automate tasks efficiently, saving time and reducing operational costs while ensuring these models remain reliable even under high task loads.

Researchers have created a headphone prototype that allows listeners to hear people speaking within a bubble with a programmable radius of 3 to 6 feet. Voices and sounds outside the bubble are quieted an average of 49 decibels, even if they're louder than those in the bubble.

Hospitals are already deploying artificial intelligence (AI) to enhance patient care. But can it actually improve doctors' diagnoses? A new study has surprising answers.

Linguistics and computer science researchers have discovered some of the root causes of why AI large language models perform poorly in human-like conversations.

Researchers have developed PanoRadar, a new tool to give robots superhuman vision by transforming simple radio waves into detailed, 3D views of the environment.

Study findings could serve as a foundation for broader neuro-monitoring applications across intensive care units globally.