Articles tagged with 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 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.
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.
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 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.
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.
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.
Researchers found that deep neural networks often respond the same way to images with no resemblance to the target, generating unnatural signals. The models develop unique invariances that are different from human perceptual systems, causing them to perceive pairs of stimuli as similar despite their differences.
Researchers investigated the role of estrogen receptor beta-positive neurons in the medial amygdala, a region involved in social information processing. They found that MeA-ERβ+ neurons exhibit different roles for receptivity-based and sex-based preferences.
Researchers created the world's largest primate brain-wide atlas using single-cell technologies, revealing over 4 million cellular profiles. The study provides a comprehensive multimodal molecular atlas to explore links between molecules, cells, brain function and disease.
Researchers have discovered that traumatic memories create new neural networks and associations between distinct networks, enabling fear-based learning and recall. The study used optical and machine-learning approaches to visualize the dynamics of brain activity during memory formation.
Researchers developed an easy-to-use optical chip that can configure itself for different functions, enabling optical neural network applications. The chip achieves positive real-valued matrix computation and demonstrates optical routing, low-loss light energy splitting, and matrix computations.
Researchers found that the medial septum, a brain area in the center of the forebrain, plays a key role in encoding and retrieval processes. It helps generate gamma oscillations, which are faster for storing and slower for recalling memories.
A team of New York University computer scientists created a neural network that can explain how it reaches its predictions, shedding light on the intricacies of RNA splicing. The breakthrough reveals how a small, hairpin-like structure in RNA can decrease splicing and provides new insights into the transfer of genomic information.
Studies suggest that humans process small and large numbers differently, with smaller numbers detected more quickly and accurately. Neurons in the brain fire at different rates for specific numbers, with errors increasing as numbers grow larger.
Researchers found that fly brain uses a three-step computation to distinguish motion patterns, dividing the workload across multiple levels. This approach helps flies detect even slight changes in motion and stay on course.
Researchers have created the first-ever map of a single animal's early visual system, reconstructing it from the eyes to neurons in a tiny parasitic wasp. The study reveals complex behaviors such as flight in an insect with just 8,600 cells compared to the human brain's 171 billion.
Researchers found that brain circuits in 'deep-blind' zebrafish are fully functional and can drive normal visual behavior through direct stimulation. This study challenges the long-held assumption that neural development depends on visual experience.
Researchers analyzed over 2 million cells from 400 postmortem brain samples to identify cellular pathways that could become new drug targets for Alzheimer's treatments. They found impairments in mitochondrial function, synaptic signaling, and protein complexes, as well as disrupted lipid metabolism.
Researchers found differential activation patterns of the Default Mode Network nodes, with facial movement inducing an inverted positive BOLD pattern in certain areas. Facial recognition and movement uniquely impact brain activity, highlighting their importance in cognitive and social development.
Researchers at Brigham and Women's Hospital have identified a common brain network associated with substance use disorder, regardless of the substance used. The finding could lead to targeted neurostimulation therapies for addiction, such as transcranial magnetic stimulation.
Scientists at the Max Planck Institute present a method for training artificial intelligence using physical processes, reducing energy consumption and computing time. The new approach relies on non-linear processes, such as optics, to mimic the human brain's parallel processing, potentially leading to more efficient neural networks.
Researchers uncover clues about how chemicals released by brain cells regulate our attention span, finding that two neurotransmitters work together in a precise sequence to regulate signal transmission. This discovery could lead to new treatments for neurological conditions associated with concentration difficulties.
Researchers discovered that brain nerve networks are organized into interconnected modules to segregate and integrate inputs, enabling efficient processing. This modular architecture allows the brain to balance local activity with global integration, essential for information representation.
Monash University researchers successfully print 3D nerve networks with living neurons that can transmit and respond to nerve signals. The networks mimic the arrangement of grey matter and white matter in the brain, allowing for significant advancements in studying nerve growth and disease mechanisms.
A new approach for coupling different light modes enables unprecedented data transfer rates in an MDM system. By using a gradient-index metamaterial waveguide, researchers achieved a high coupling coefficient and created a 16-channel MDM communication system with a data transfer rate of 2.162 Tbit/s.
A UMass Amherst neuroscientist is mapping the brain of a sea slug to study how neurons are added to functional neural circuits, shedding light on how this process contributes to neurological conditions. The project aims to provide an unprecedented look at brain development and potentially inform human brain development.
A new study led by Dr. Armen Saghatelyan uncovered the migratory mechanisms of neuronal cells in a neurodevelopmental disorder. The team found that modulating autophagy with FDA-approved drug metformin restored the cells' migratory properties.
Researchers developed a high-performance photonic spiking neural network that surpasses traditional digital systems with its ultrafast performance and low power consumption. The new network achieved excellent classification accuracies of over 94%, outperforming benchmark results with small training sets.
Researchers aim to understand the role of gaze in post-stroke reading impairments and develop an employment-specific assessment tool for individuals with autism. The studies may provide major steps towards finding innovative solutions for individuals affected by neglect dyslexia and autism.
Researchers used a new analytical approach to understand how the brain controls movement and eye stability in zebrafish. By analyzing neuronal activity, they identified two main features that correspond to specific types of movements, including eye rotation and body positioning.
Researchers have identified a conserved brain region in vertebrates essential for initiating forward movement. The mesencephalic locomotor region (MLR) stimulates reticulospinal neurons to control finer details of movement. Mapping this circuit could lead to new treatments for Parkinson's disease.
A paper published in Nature Communications shows that when neurons are given task-related sensory input, they change how they behave and support the critical brain hypothesis theory. The researchers used DishBrain, a collection of 800,000 human neural cells, to demonstrate near-critical network behavior emerges when the neural network ...
Researchers found that insects like American cockroaches use the mushroom body to encode behavioral decision-making based on sensory information. The study challenges the prevailing view of insect cognition, suggesting a more complex brain function than previously thought.
A new study by North Carolina State University found that artificial intelligence performs better when it chooses diversity over lack of diversity. The AI was able to increase its accuracy up to 10 times more than conventional AI in solving complicated problems.
Researchers have developed a novel neural network approach to design brand new proteins with unique arrangements and dynamic functionalities. The method combines attention neural networks with graph neural networks to predict existing protein properties and envision new proteins that nature has not yet devised.
A hybrid system of electronic encoding and diffractive optical decoding transmits optical information with high fidelity through random, unknown diffusers. The system outperforms traditional approaches that only utilize a diffractive optical network or an electronic neural network for optical information transfer.
Researchers use surface normal nonlinear photodetector to improve speed and energy efficiency of diffractive optical neural networks. The new device can perform high-speed image and video processing at the speed of light in an extremely energy efficient manner.
A study found that dogs are more sensitive to speech directed at them than adult-directed speech, with a greater response to female voices. The researchers measured dog brain activity via fMRI and found that the sensitivity was affected by voice pitch and variation.
A recent study published in Neuron reveals that Noelin proteins play a crucial role in learning and memory formation in the mammalian brain. The study found that these proteins act as 'universal anchors' controlling the distribution and dynamics of AMPA receptors, which are essential for synaptic plasticity.
Researchers propose a new approach to reconfigurable diffractive neural networks using metasurfaces, enabling flexible AI computing for various tasks. The proposed pluggable diffractive neural networks demonstrate high classification accuracy and efficiency, with potential applications in real-time object detection and intelligent opti...
Researchers propose a hypothesis that astrocytes, non-neuronal cells in the brain, can perform core computation as transformers, providing insights into human brain function and machine learning success. This discovery could spark future neuroscience research and help explain transformer performance across complex tasks.
Researchers discovered intricate neural connections in Octopus vulgaris, a model organism for studying memory acquisition networks. The findings challenge traditional models of neural network functionality, revealing an evolutionary adaptation that underlies the octopus's unique cognitive prowess.
Researchers found an abnormal imbalance of excitatory cortical neurons in people with autism spectrum disorder, depending on their head size. The study used human 'mini-brain' models called organoids to recreate the brain development alteration that occurred in patients during fetal development.
A recent study at the University of Helsinki found that speech production and singing are supported by the same neural network in the brain, challenging previous notions about their separate functions. The left hemisphere plays a crucial role in both abilities, particularly in terms of singing.