Articles tagged with Neural Networks
A new computational model uses neural network architecture to generate accurate predictions of airflow while reducing computational cost. It can capture most of the original physics in a flow prediction with relatively little processing complexity.
The new AI model uses a visual map to explain each diagnosis, helping doctors follow its line of reasoning and check for accuracy. The tool aims to catch diseases in their earliest stages, making it easier on doctors and patients alike.
Researchers have developed a non-crossing quantile regression neural network (NCQRNN) model to enhance the statistical reliability of weather forecasts. The NCQRNN model preserves the rank order of output nodes, ensuring lower quantiles stay smaller than higher ones, boosting accuracy and improving forecast interpretability.
A new method called NIRE enables large-scale and long-term observation of neuronal structures and activities in awake mice. The method uses fluoropolymer nanosheets covered with light-curable resin to create larger cranial windows, allowing for high-resolution imaging with sub-micrometer resolution.
UCLA Health researchers found a group of astrocytes in the central striatum regulate neurotransmitter communication and gate perseverative behavior. The discovery may lead to potential therapies for disorders like autism, OCD, and Tourette syndrome.
Dr. Ryyna Ethell's lab at UC Riverside has been awarded a $2.4 million NIH grant to investigate the role of astrocytes in inhibitory synapse development and their connection to neurodevelopmental disorders like ADHD and autism.
A new AI model developed by MIT researchers breaks down complex warehouse navigation into smaller chunks, identifying optimal areas for decongesting robots. The technique improves efficiency by nearly four times, opening up potential applications in other complex planning tasks.
A team of researchers used deep brain stimulation to localize disrupted neural pathways in patients with Parkinson's disease, dystonia, OCD, and Tourette's syndrome. The study identified specific brain circuits associated with each disorder, revealing overlapping malfunctions that suggest a complex network of brain dysfunctions.
Researchers developed artificial neural networks using silicon microresonators, offering a promising platform for efficient AI systems. The devices mimic biological neurons' nonlinear behavior, enabling precise control of light properties.
A new mouse study has uncovered the neural connections that help shape decisions, finding sequential groups of neurons activated and suppressed to reinforce a choice. The research combines structural, functional, and behavioral analyses to explore how neuron-to-neuron connections support decision-making in the brain.
Researchers at the University of Pennsylvania have developed a new silicon-photonic chip that can perform vector-matrix multiplication using light waves, allowing for accelerated AI computing. The chip's design has privacy advantages, as sensitive information can be processed simultaneously without being stored in memory.
Researchers have identified a network of neurons controlling right-left movements in the brain, which may help treat Parkinson's disease. The discovery provides insight into how essential movements are produced by the brain.
Researchers have discovered that Schwann cells, previously thought to be solely responsible for protecting nerve fibers, also detect sensory stimuli such as touch, heat, and cold. The findings open new avenues for understanding and treating pain and impaired touch perception.
A deep learning model predicts Antarctic sea ice will remain near historical lows in 2024, with a slight increase in extent compared to the record low in 2023. The model accurately reproduced three historical summer lows and showed promise for seasonal prediction of Antarctic sea ice.
A study of over 580 children found that childhood trauma disrupts neural networks involved in self-focus and problem-solving. The research suggests that trauma therapies should address not only thoughts but also the impact on the body, sense of self, emotional processing, and relationships.
Researchers trained an AI model on a single child's headcam video recordings from six months to two years, finding it could learn substantial words and concepts. The model linked words to visual counterparts, generalizing them to different instances, reflecting children's lab-based word learning.
A study of 205 adolescent football players and 70 noncontact control athletes reveals discernible structural and physiological differences in brain regions associated with mental health. These findings suggest a potential link between football playing and mental health outcomes.
Researchers at UW-Madison have developed a groundbreaking method for 3D printing functional human brain tissue, which can grow and function like typical brain tissue. The printed cells form connections, send signals, and interact with each other through neurotransmitters, mimicking the complexity of human brains.
A newly developed neural network is highly accurate in identifying key landmarks important in breast surgery, enabling rapid and automated detection of breast symmetry. The network's performance was tested using a set of photographs of patients who underwent breast reconstruction after cancer surgery.
A KAIST research team led by Professor Hawoong Jung identified the principle behind musical instincts emerging from the human brain without special learning using an artificial neural network model. The study found that cognitive functions for music forms spontaneously as a result of processing auditory information received from nature.
Researchers have designed a new, affordable system to study neural interactions and compute using living neurons. The open-source MiV system boasts over 500 electrodes, offering improved control and precision in measuring neural processes.
Researchers extend spatially incoherent diffractive networks to perform complex-valued linear transformations with negligible error, opening up new applications in fields like autonomous vehicles. This breakthrough enables the encryption and decryption of complex-valued images using spatially incoherent diffractive networks.
A new study using generative AI models simulated how the brain learns and remembers events, revealing how memories are re-constructed in our minds. The model showed how the hippocampus and neocortex work together to create efficient 'conceptual' representations of scenes, enabling us to both recall past experiences and imagine new ones.
Researchers analyzed large datasets of neural wiring in fruit flies, mice, and worms to compare networks across species. They created a mathematical model based on Hebbian plasticity, which shows how strong connections form and leads to clustering, suggesting a shared principle of self-organization governs brain network formation.
Researchers propose a simple model that accurately describes neuronal connectivity in various organisms, suggesting that general networking principles govern brain organization. The model also provides an unexpected explanation for clustering phenomenon in social interactions and can be extended to other types of networks.
Researchers introduce a cycle-consistency-based uncertainty quantification technique to enhance the reliability of deep neural networks in solving inverse imaging problems. The method uses forward-backward cycles to estimate network uncertainty, demonstrating improved accuracy in detecting image corruption and out-of-distribution images.
A new study led by Dr. Richard Naud of the University of Ottawa's Faculty of Medicine tackles the mystery of neuronal response variability, controlling output with dendrites' inputs to the core and little antennas
A new transparent brain implant has been developed to read deep neural activity from the surface, providing a step closer to building a minimally invasive brain-computer interface. The technology enables high-resolution data about deep neural activity by using recordings from the brain surface and correlating them with calcium spikes i...
Researchers investigated role of estrogen receptors in lateral septum regulating social anxiety in male mice, finding that ERβ plays a crucial role in neural circuitry. Knockdown effects on ERβ gene expression showed increased social anxiety, highlighting its importance in social situations.
A novel human brain organoid model generates all major cell types of the cerebellum, including functional Purkinje neurons. This breakthrough provides a new way to explore cerebellar development and disorders, advancing therapeutic interventions.
Researchers discovered that microglia regulate neuronal activity in a brain region-specific manner, playing a key role in how anesthesia works. Microglia depletion delayed induction and led to early emergence of anesthesia.
A recent study using AI to analyze registry data on people's residence, education, income, health, and working conditions can predict life events such as personality and time of death. The model outperforms other advanced neural networks and provides precise answers despite ethical concerns about sensitive data and bias.
A new study from MIT shows that computational models trained on auditory tasks display an internal organization similar to the human auditory cortex. Models trained on diverse tasks and background noise more closely mimic brain activation patterns.
A recent study reveals the importance of brain hubs in neural networks and their rapid compensation when lost. The researchers obtained direct recordings of human brain activity before and after surgically disconnecting a critical language hub.
Researchers have developed an algorithm to train an analog neural network just as accurately as a digital one, decreasing energy consumption and eliminating the need for a digital twin. This approach is more biologically plausible and shows improved speed, robustness, and reduced power consumption compared to other methods.
Researchers at University of Tsukuba found that hydrogen sulfide production within the respiratory center alters neurotransmissions, disrupting breathing patterns. The study identified variations in this mechanism across different regions, revealing a modulating influence on neural circuits contributing to respiration stability.
Scientists developed an AI method to track neurons in moving and deforming animals using convolutional neural networks with targeted augmentation. This breakthrough reduces manual annotation efforts by three times, enabling faster analysis of brain activity in model organisms like Caenorhabditis elegans.
A new study reveals AI tools are more vulnerable than thought to targeted attacks that force AI systems to make bad decisions. Researchers developed a software called QuadAttac K to test for vulnerabilities in deep neural networks.
A new study finds that exposure and response prevention (EX/RP) therapy for OCD improves cognitive control by strengthening connections between brain networks. This breakthrough sheds light on the mechanisms underlying EX/RP's effectiveness in treating OCD, paving the way for targeted therapies.
Researchers developed three diffractive deep neural networks using orbital angular momentum to recognize objects in images, achieving accuracy comparable to wavelength and polarization-based models. The technology has potential for real-time processing applications like image recognition and data-intensive tasks.
A study by the University of Ottawa found that various long-term memory types, such as semantic and episodic memory, rely on the same brain network to varying degrees. The research used neuroimaging techniques to compare brain activity across different memory conditions.
A new publication introduces a multi-synaptic photonic SNN based on a DFB-SA chip, showcasing its ability to simulate neuron-like dynamics and improve network performance. The algorithm's learning efficiency and performance were excellent, with improved robustness through time coding.
A new photonic integrated synaptic chip simulates biological neurons' plasticity and achieves nonlinear spike activation rates of 2GHz. This innovation enables the realization of monolithic large-scale photonic spiking neural networks, promoting applications in data centers, edge computing, and autonomous driving.
Researchers at Rice University are developing a machine learning framework to improve decision-making processes in military communication networks. The goal is to enable rapid, adaptive action across a broad range of scenarios by combining local data in the most effective manner.
A new tool called Facemap uses deep neural networks to relate mouse facial movements to neural activity in the brain. This allows researchers to track and quantify movements and correlate them with brain activity, bringing them one step closer to understanding how the brain uses persistent, widespread signals.
Researchers at Purdue University developed a new tool to visualize neural network decisions, making it easier to identify errors in image recognition. The tool uses graph-topological data analysis to provide a bird's-eye view of all images in a database, revealing areas where the network struggles to distinguish between classifications.
The Python code library snnTorch, developed by UC Santa Cruz's Jason Eshraghian, has surpassed 100,000 downloads and is used in various projects. A new paper published in the Proceedings of the IEEE documents the library and offers a candid educational resource for students and programmers interested in brain-inspired AI.
Researchers developed a deep convolutional neural network to pinpoint cardiac catheter tip locations in photoacoustic images, achieving high precision and recall. The approach has the potential to replace fluoroscopy during cardiac interventions, leading to safer procedures.
A recent study published in the Proceedings of the National Academy of Sciences found that AI's deep convolutional neural networks can identify faces but struggle to capture other important information like emotional state and trustworthiness. Brain activity scans revealed a weak correlation between AI's codes and human brain represent...
Researchers from the University of Cambridge have developed a virtual reality application that allows users to build figures and shapes without interacting with menus. The 'HotGestures' system uses machine learning to recognize hand gestures, providing fast and effective shortcuts for tool selection and usage.
A team of scientists has created the first comprehensive map of a worm's nervous system, showing how neurons communicate wirelessly through neuropeptides. The study reveals complex networks with distinct hubs and connections that transcend traditional synaptic pathways.
A team of neuroscientists and physicists at Princeton University studied the brain of Caenorhabditis elegans to understand how information flows through a network of interacting neurons. They used optogenetics to activate individual neurons and observe how other neurons responded, shedding light on the complex neural connections.
Researchers are combining biology, physics, computer science, and engineering to design electric circuits that mimic the brain's adaptive behavior. The goal is to create a more efficient AI application that can learn from history and adapt without significant energy consumption.
A new hybrid method developed by Concordia researchers combines data from Weibull probability distribution and numerical weather prediction models to improve wind speed forecasting accuracy. This innovation has the potential to significantly enhance urban power generation, particularly in areas with high variability in wind speeds.
A new method called TWC-Swin effectively restores holographic images even under low spatial coherence and arbitrary turbulence, surpassing traditional convolutional network-based methods. The study demonstrates strong generalization capabilities, extending its application to unseen scenes.
A new parallel hybrid quantum neural network demonstrates improved performance by combining the strengths of both quantum and classical layers. The model outperforms traditional machine learning methods in processing complicated patterns and relationships from data inputs.
Researchers at New York University developed a novel learning procedure called Meta-learning for Compositionality (MLC) that enables neural networks to make compositional generalizations. MLC outperforms existing approaches and is on par with, and in some cases better than, human performance.
Researchers at Sainsbury Wellcome Centre find frontal and parietal cortex play key role in encoding value of economic choices when faced with uncertainty. The study provides foundation for understanding neurobiology of risky decisions.
A team of researchers proposes a novel approach to generate three-dimensional holograms directly from regular 2D color images captured using ordinary cameras. This approach utilizes deep learning to transform the image into data that can be used to display a 3D scene or object as a hologram.
A team of engineers at the University of Pittsburgh has created a new robot that can navigate complex environments using bio-inspired neural networks. The robots will be able to avoid obstacles and learn from their experiences, enabling autonomous navigation for various applications such as commercial transport and disaster response.