Articles tagged with Neural Networks
The University of Essex team has devised a new approach to training neural networks called Target Space, which stabilizes the learning process by tweaking neuron firing strengths. This method enables deeper neural networks with fewer training examples and computing resources, accelerating AI breakthroughs.
Research reveals that the brain's electrical rhythms fluctuate between high precision and low precision states several times per second, affecting how relevant information is transmitted. Cross-frequency coupling enables selective attention by modulating the strength of different frequencies, while distinguishing between different type...
Researchers studied how diverse neural network training datasets impact generalization. They found that data diversity is key to overcoming bias, but also degrade performance when neural networks are trained for multiple tasks simultaneously. The study highlights the importance of designing diverse and controlled datasets in machine le...
Researchers at Tokyo Institute of Technology have developed a new AI processor called Hiddenite, which achieves state-of-the-art accuracy in sparse neural networks with lower computational burdens. The chip drastically reduces external memory access for enhanced computational efficiency.
A team of researchers from Skoltech and universities developed a neural network-based solution for automated recognition of chemical formulas on research paper scans. The algorithm combines molecules, functional groups, fonts, styles, and printing defects to mimic existing molecular template depiction styles.
A new machine learning process has been developed to identify and classify hip fractures from X-rays, achieving 92% accuracy and confidence. This approach aims to improve patient outcomes, reduce care costs, and alleviate the bottleneck of unreported radiographs in the UK.
Neuroscientists have designed neural organoids with both mature neurons and astrocytic glial cells to study interactions between brain cells. The new technology enables the emulation of brain activity during healthy and disease states, opening doors to rapid drug screening for neurological diseases.
Researchers propose a new neural network-based method to visualize chemical reactions in a 2D plane, grouping similar reactions together. The visualization helps chemists understand the global chemical reaction space and identify underused or unused reaction types.
Researchers discovered increased activity in the hippocampus during anesthesia and sleep, preceding Alzheimer's symptoms by years. This abnormality may enable early diagnosis and treatment of the disease.
Researchers at KTH Royal Institute of Technology and Stanford University have developed a material that enables the commercial viability of neuromorphic computers mimicking the human brain. The material, MXene, combines high speed, temperature stability, and integration compatibility in a single device.
The MIT team developed a computer model that can perform sound localization tasks as well as humans, and adapts to real-world environments. The model uses convolutional neural networks and was trained on over 400 sounds, including human voices and animal sounds.
Cornell researchers have successfully trained various physical systems, including mechanical, optical, and electrical systems, to perform machine learning tasks. The developed training algorithm enables diverse systems to be chained together for efficient processing.
Scientists at Vienna University of Technology have developed a new type of neural network that can accurately simulate the quark-gluon plasma, a state of matter present in the early universe. The networks use gauge invariant convolutional neural networks to recognize patterns and predict properties of the plasma.
Researchers found that REM sleep has specific properties allowing rapid arousal in response to predatory stimuli. The medial subthalamic nucleus (mSTN) plays a key role in this process, producing a lowered arousal threshold during REM sleep for detecting predator threats.
MIT researchers develop a method to test feature-attribution methods for machine-learning models. They find that even the most popular methods often miss important features in an image and some perform as poorly as a random baseline. This has major implications for high-stakes situations like medical diagnoses.
A team of scientists from Gwangju Institute of Science and Technology developed a deep learning-based approach to predict SC2 battle outcomes by considering army composition and terrain type. The proposed model leveraged parameter sharing, enabling it to analyze complex factors accurately and make predictions.
A new project aims to improve the performance of Graph Neural Networks (GNNs) by leveraging weak supervision and additional information. The research has potential applications in fraud detection, agriculture, and cancer diagnosis.
Researchers at RIKEN CBS demonstrate that neural networks minimize energy cost and solve mazes efficiently, pointing to a set of universal mathematical rules. The findings will aid in analyzing impaired brain function and generating optimized neural networks for artificial intelligences.
A two-year study aims to restore functional vision among veterans and service members with head trauma-related visual dysfunction using immersive virtual reality technology. Researchers will explore the feasibility and benefits of applying new technologies to rehabilitative treatment activities.
A team of researchers used neural network data to study grid cell activity in the brain, finding that collective neural activity is shaped like a torus, or doughnut. The study provides new insights into how large networks of neurons produce properties that cannot be inferred from individual cells.
Researchers have discovered a new biomarker, microstate D, associated with increased sleep disturbance and inattention symptoms in adults with ADHD. The study, published in Biological Psychiatry: Cognitive Neuroscience and Neuroimaging, provides evidence for a more precise diagnosis of the disorder.
A transdisciplinary research team at Göttingen Campus has found a new perspective on the rhythmic processes in the brain. They discovered that adapting interneurons can switch between very slow rhythms and fast rhythms, challenging previous assumptions about their function.
A Salk Institute team has uncovered a neural network in the brain that connects breathing rhythm with feelings of pain and fear. This discovery could lead to the development of an analgesic that prevents opioid-induced respiratory depression, a major cause of overdose deaths.
A team of researchers from the University of Groningen developed an AI-based system that can identify individual Holstein cows in a milking station based on their coat pattern. The system achieved a recognition rate of 99.7% and has several advantages, including non-invasiveness, cost-effectiveness, and scalability.
Researchers at the University of Göttingen studied how blocking certain enzymes affects brain adaptability in healthy and diseased mice. In healthy mice, inhibiting these enzymes blocked neuronal plasticity, while in stroke-affected mice, it restored lost plasticity.
Washington University researchers have designed a new processing-in-memory (PIM) circuit that can increase PIM computing's performance by orders of magnitude. The circuit uses resistive random-access memory PIM, allowing for analog computations and eliminating the need for digital conversions.
Researchers aim to develop AI agents that reuse information, adapt quickly to new conditions and collaborate by sharing experiences. The goal is to enable machines to continually learn from their collective experiences and improve performance on novel and previous tasks.
A Michigan Tech-developed machine learning model uses probability to classify breast cancer shown in histopathology images and evaluate the uncertainty of its predictions. The model outperforms similar models and can measure uncertainty, promising time savings and referrals to human experts.
A new strategy for ultrafast and energy-efficient all-optical computing is proposed based on convolutional neural networks. The approach uses cascaded silicon waveguides to control light amplitude and phase, achieving ultrafast computing times of several picoseconds with low energy consumption.
Researchers at Tohoku University and the University of Gothenburg have developed a new spintronic technology that integrates a memristor-controlled oscillator array, allowing for efficient brain-inspired computing. This breakthrough enables sophisticated cognitive tasks like image recognition with reduced energy consumption.
Researchers at the University of Gothenburg have successfully combined a memory function with a calculation function in the same component, enabling more efficient technologies like mobile phones and self-driving cars. The discovery opens the way for brain-like computers that can perform tasks effectively and energy efficiently.
Researchers discovered the retrosplenial cortex as the site of value decision-making in the brain. Persistency allows value signals to be effectively represented across different brain areas, especially the RSC. Artificial intelligence networks mimicking mouse decisions showed remarkably similar results.
A KAUST team developed an improved method for detecting malicious intrusions using deep learning, achieving accuracy rates of up to 99% in simulations of different kinds of attacks. This stacked deep learning approach promises an effective defense against cyberattacks and could prevent outages in critical infrastructure.
A study has developed a method using dark-field microscopy and deep learning algorithms to identify microplastics in human cells, achieving an accuracy of 93% for 1-micron polystyrene particles. The technique has the potential to screen microplastics in various samples, reducing time-consuming data acquisition and processing steps.
A new study by Ritsumeikan University reveals that whole-body dynamic balance training increases brain connectivity and improves motor learning. The research found significant changes in brain activity associated with offline learning, opening up new avenues for studying neural networks during natural tasks.
A new AI model has been developed to classify colorectal polyps, demonstrating accuracy and sensitivity at a level comparable to practicing pathologists. The model was tested in a clinical trial involving 15 pathologists, showing significant improvements in accuracy compared to traditional methods.
Researchers found that re-identifying individuals from genomic data using public face images is harder than previously thought, with success rates well below idealized settings. They developed a method to alter social media photos and reduce the risk of privacy breaches.
Researchers at Japan Advanced Institute of Science and Technology create method to estimate five-room acoustic parameters and speech transmission index using a short conversation, with potential applications in monitoring auditoriums during concerts and saving lives through smart speakers
The research team developed a technology for remotely assessing the condition of a building during an earthquake based on the readings from the building's seismometer. The new method uses CNN machine learning to quickly assess damage levels and determine if a building can continue to be used.
A study involving Drosophila found that a constant and precisely regulated energy supply is essential for nerve development, particularly during the degradation of nerve connections. Malnutrition was shown to intensify defects in this process.
Convolutional neural networks trained to identify abnormalities on upper extremity radiographs are susceptible to a ubiquitous confounding image feature: radiograph labels. Covering these labels increases accuracy, while using them alone leads to decreased performance.
Research suggests that larger bee colonies with comfortable food stores are less willing to take risks, while smaller colonies with limited resources are more likely to ignore warning signals. This study provides insights into the complex communication system of bees and its implications for understanding biological collectives.
A study at the University of Turku found that brain function regulating satiety and appetite is altered before obesity develops, with family background risk factors contributing to these changes. The findings suggest the brain and central nervous system are key targets for treating obesity.
Researchers at the University of Waterloo have created a deep neural network that detects disease biomarkers with high accuracy, achieving 98 per cent detection of peptide features. This breakthrough could enable earlier and more accurate disease detection through tissue sample analysis.
Scientists at Gladstone Institutes discovered that non-convulsive epileptic activity drives chronic brain inflammation in Alzheimer's models, which can be reversed by eliminating protein tau or using the anti-epileptic drug levetiracetam. This link between brain networks and immune cells may hold promising treatments for Alzheimer's di...
Researchers have trained a neural network to detect anomalies in medical images, adapting it to the nature of medical imaging and achieving better results. The new method uses weakly supervised training and can spot small-scale anomalies, accelerating the work of histopathologists and radiologists.
Scientists studied rats in a virtual reality maze to gain a deeper understanding of the hippocampus' circuit-level functions. They discovered that hippocampal neurons encoded multiple aspects of an animal's location, including distance traveled and body direction.
A pilot study examines graph-theoretical properties of brain networks in traumatic brain injury and controls, showing association with balance impairment and structural damage. The study uses EEG-based functional connectivity measures during a balance perturbation task to explore underlying neural mechanisms.
Researchers at Gladstone Institutes found that reducing tau levels impacts both excitatory and inhibitory cells, leading to a reduction in excitation-inhibition ratios. This effect counteracts diseases that cause abnormal increases in this ratio, potentially improving the brain's ability to perform its functions.
Researchers developed an attention-based deep neural network to detect multiple ship targets, exceeding conventional networks' performance. The model focused on inherent features of the two ships simultaneously, outperforming traditional approaches.
A novel machine learning approach has been developed to understand symmetry and trends in materials, enabling researchers to group similar classes of material together. The technique uses a large, unstructured dataset gleaned from 25,000 images to identify structural similarities and trends.
Researchers have made progress in creating a brain atlas of the mouse brain, which will help develop tools for studying the human brain. The study describes the diversity of neurons in the mouse brain and establishes new methods for characterizing cell types and neural connections.
Researchers at the Sainsbury Wellcome Centre have discovered a new brain circuit that enables mice to override their instincts based on previous experience. The ventral lateral geniculate nucleus (vLGN) inhibits threat reactions when animals feel safe, but activates them when danger is perceived.
Researchers used fractal analysis to study brain network patterns while listening to a story. The results show that complex thoughts are reflected in high-order dynamic correlations in neural activity patterns.
Researchers developed a novel approach to 3D image segmentation, segmenting the gaps between parts instead of contours, to automate tedious tasks. The technique demonstrates promising results in diagnosing TMJ-related issues and has potential applications in other fields.
A team of researchers has found a 390-million-year-old hyper-facet eye system in trilobites that is unique to the animal kingdom. The discovery suggests that this ancient eye may have been an adaptation for life in low light conditions, and could provide insights into the evolution of visual systems.
Scientists discovered that recurrent neural networks (RNNs) play a crucial role in the frontal cortex, responsible for decision-making, expressive language, and voluntary movement. The research also found that RNNs are more complex than previously thought, with a unidirectional structure.
Scientists at Aarhus University are working on a nano-sized brain-inspired computer that can harvest its own energy, making it the smallest and most efficient AI system yet. The project aims to reduce power consumption by 12 orders of magnitude compared to modern supercomputers.
Researchers have developed a new microscopy technique, light beads microscopy, that captures detailed images of activity from one million neurons across the mouse brain at high speed and single-cell resolution. This innovation allows scientists to investigate biological questions in a way that was not possible before.
Researchers at University of Chicago and Argonne National Laboratory found that primate neurons receive two to five times fewer excitatory and inhibitory synaptic connections than similar mouse neurons. The study suggests that metabolic costs may drive larger neural networks to be sparser, as seen in primates versus mouse neurons.