Articles tagged with Image Processing
A team of researchers from The University of Osaka has developed a new approach for depth reconstruction from defocus, estimating distances by analyzing blur in an image. Their method combines a coded-aperture camera with diffusion-model-based AI to accurately estimate depth and produce high-quality images.
FireANTs, an open-source algorithm, combines AI optimization and geometry to quickly match complex medical images. The new method can accomplish what took weeks in minutes, detecting subtle changes that signal disease or cognitive decline, making it practical for clinical practice.
A new ultra-lightweight AI model, Multinex, advances low-light image enhancement by leveraging classical colour vision theory and Retinex principles. The model outperforms comparable compact systems, recovering detail and clarity from previously unusable images.
Researchers from Drexel University developed BioCoach, a program using AI and computer vision to analyze video and provide form coaching in real time. The system analyzes visual appearance and motion patterns, as well as 3D skeletal movements and body shape, to deliver detailed biomechanics-based feedback.
Researchers developed a disco laser system to enhance data visualization for snow groomers, improving operator comfort and reducing nausea caused by VR headsets. The system also enables better tracking and orientation aids, leading to more efficient and safe operation in challenging conditions.
Scientists successfully built the smallest X-ray interferometer to measure how X-rays interact with atomic nuclei. This breakthrough technology enables precise measurement of X-ray refraction and provides new avenues for research.
A team of scientists from NTU Singapore has developed a new biochip that, when paired with Artificial Intelligence (AI), can detect quickly and accurately extremely small amounts of microRNAs. The device can cut detection time from hours to 20 minutes.
Researchers developed a dynamic range compression dual-domain attention network to tackle extreme exposure conditions in tunnels. The DRC-DFANet model optimizes global illumination coordination and local detail restoration, preserving fine details while adjusting brightness intelligently.
Researchers develop a comprehensive framework for parallel single-pixel imaging, enabling accurate separation and localization of complex illumination components. The proposed model shows advantages in mixed-scene reconstruction and significantly outperforms conventional methods in low signal-to-noise ratio environments.
HeapGrasp uses RGB images to analyze object silhouettes and estimate its 3D shape, reducing the need for depth information. The approach achieves high accuracy while minimizing camera movement and execution time.
Using AI-driven analytical methods, researchers have created a custom OCT system that enables the objective measurement of wound progress over time. The platform shows that stiffer mechanical properties improve wound healing outcomes, with faster transition to intact regenerated tissue.
A blind patient partially recovered natural vision through electrical stimulation of the visual cortex, independent of the implant. The recovery was sustained over time and remained even after the device was removed, suggesting individual factors may have contributed to this outcome.
Researchers created a new frequency-aware approach to crafting adversarial images that better match human visual perception. The method, called Input-Frequency Adaptive Adversarial Perturbation (IFAP), significantly outperformed existing techniques in structural and textural similarity.
Researchers developed AI-powered in silico labeling to analyze cell images without staining, preserving cell health. The system leverages context, such as cell shape and position, to accurately stain rare processes like cell division.
A new AI tool, CattleFever, uses artificial intelligence and thermal cameras to estimate cattle body temperature from a photo. The system can automatically determine an animal's body temperature within 1 degree of the reading from a thermometer.
A reconfigurable optical computing platform based on a double-layer liquid crystal structure has been developed to enable multifunctional all-optical image processing. The platform integrates eight types of image processing functions in one go, including bright-field imaging, vortex filtering and edge enhancement, promising substantial...
Researchers found that AIs consistently converged on 12 common themes despite diverse prompts, suggesting biases in training data. The models failed to generate novel or creative outputs, highlighting the need for anti-convergence mechanisms and human input for AI's creative potential.
Researchers have launched a new multimodal medical dataset, Observer, capturing anonymized, real-time interactions between patients and clinicians. The dataset links video, audio, transcripts, and electronic health records to study subtleties like body language and environmental factors affecting care.
The Association for Computing Machinery has named 61 new Distinguished Members, recognized for their contributions to AI for healthcare, data management, human-computer interaction, and other areas. The recipients include experts from top universities, corporations, and research institutions worldwide.
Researchers explore ways to understand and control multiple errors affecting machine tool accuracy, combining traditional models with data-driven approaches and digital twin technology. This enables more integrated systems that can monitor themselves, predict changes, and adjust behavior automatically.
The Center for Computational and AI-enabled Imaging Sciences brings together experts to develop AI-powered medical imaging applications that integrate information from different imaging types. This may include identifying previously unknown early indicators of disease onset.
Researchers at the University of Cambridge found that the human eye has a resolution limit, above which screens provide more information than can be detected. For an average UK living room, a 44-inch 4K or 8K TV does not offer additional benefits over a lower resolution Quad HD TV.
Researchers developed AI-powered BlinkWise glasses that track blinking patterns to assess fatigue, mental workload, and eye-related health issues. The device uses radio signals to detect minute eyelid movements with unprecedented detail, preserving privacy and using minimal power.
A novel detection framework, SORA-DET, is introduced for UAV remote sensing, achieving high accuracy and efficiency while being compact and fast. It outperforms large-scale models with up to 88.1% fewer parameters.
Researchers developed MoBluRF, a two-stage motion deblurring method for NeRFs, achieving high-quality 3D reconstructions from ordinary blurry videos. The framework outperforms state-of-the-art methods and is robust against varying degrees of blur, enabling smartphones to produce sharper and more immersive content.
Astronomers have developed a protocol to detect supernovae within 24 hours of their explosion, using high-cadence sky surveys. The method involves rapid searches for candidates based on light signal absence and galaxy location, followed by spectroscopic observations to determine the type of supernova.
Researchers developed a novel approach called R3DG that analyzes representations at varying granularities to capture nuanced emotional fluctuations and reduce computational complexity. This framework demonstrates superior performance in multiple multimodal tasks, including sentiment analysis, emotion recognition, and humor detection.
A research team developed an innovative unsupervised model for industrial anomaly detection using paired well-lit and low-light images. The model leverages feature maps, Low-pass Feature Enhancement, and Illumination-aware Feature Enhancement to detect anomalies while remaining lightweight and memory-efficient.
A new design method, SPADE, evaluates UAV video surveillance system quality without real-world manipulations. It assesses the impact of image resolution on object detection performance., The SPADE method can be applied to various UAV systems and will aid in designing complex large-scale systems involving multiple UAVs.
Researchers have demonstrated a new technique, RisingAttacK, to manipulate all widely used AI computer vision systems, allowing them to control what the AI 'sees'. The attack is effective at influencing the AI's ability to detect top targets, such as cars, pedestrians, or stop signs.
A cardiac magnetic resonance imaging (MRI) scan usually takes anywhere from 30 to 90 minutes. The AI-assisted model developed by Mizzou researchers can turn low-quality MRI heart scans into high-quality images in less time, while improving patient experience and reducing costs.
A new study reveals a five-fold increase in computer vision papers linked to surveillance patents, highlighting the rise of obfuscating language that normalises surveillance. The top institutions producing surveillance are Microsoft, Carnegie Mellon University, and MIT.
Researchers at KAIST have developed a technology to enhance creative generation of AI generative models like Stable Diffusion, generating novel and useful images. The algorithm amplifies internal feature maps to boost creativity without new training, outperforming existing methods in novelty and utility.
Researchers at ETH Zurich have developed a novel solution for image sensors, utilizing lead halide perovskite to capture every photon of light. This allows for improved color recognition and higher resolution, as well as advantages in hyperspectral imaging.
Scientists have developed a groundbreaking adaptive optics system that removes blur from images of the Sun's corona, revealing clearest images to date. The technology has produced remarkable observations of fine-structure in the corona, including raindrops and turbulent internal flows.
University of Missouri researchers create digital sentiment map using AI to analyze public Instagram posts, linking emotional tone to real-life features. The tool aims to improve city services, identify areas of concern, and inform emergency response decisions.
Researchers have developed CaliAli, an advanced analytical framework that aligns calcium imaging data across multiple sessions, allowing for the continuous tracking of individual neurons. This breakthrough enables long-term brain activity studies and advances understanding of memory formation, retention, and neurological diseases.
Researchers at Pohang University of Science & Technology have developed Pixel-Based Local Sound OLED technology, allowing each pixel to emit different sounds. This breakthrough enables truly localized sound experiences in displays, enhancing realism and immersion.
Researchers at Pohang University of Science & Technology (POSTECH) have developed an achromatic metagrating that handles all colors in a single glass layer, eliminating the need for multiple layers. This breakthrough enables vivid full-color images using a 500-µm-thick single-layer waveguide.
Recent high-quality deepfake videos can feature realistic heartbeats and minute changes in face color, making them challenging to detect. Researchers found that even small variations in skin tone and facial motion can replicate the original pulse in deepfakes.
A team of researchers from Jinan University has developed a metasurface-based imaging technique that can precisely measure the intensity, phase, and polarization of arbitrary light field distributions in a single exposure. The system uses optimized metasurface structural parameters to diffract incident light fields into sub-images that...
Researchers at Johns Hopkins University found that AI systems struggle to understand social dynamics and context necessary for human interaction. Human participants were able to accurately rate features important for understanding social interactions, while AI models failed to match human brain and behavior responses across the board.
A collaborative research team has developed a novel mixed reality (MR) technology that uses real-world doors as natural transition points. The system allows users to select a door within their MR interface and seamlessly transition into a virtual space, creating an unprecedented sense of immersion.
Researchers have developed a novel method, PEDL, to improve the performance of photoacoustic microscopy. PEDL seamlessly integrates physical principles and domain-specific knowledge into a deep learning model, enabling accurate simulation of physical processes in PAM. This leads to enhanced image resolution and improved accuracy in dee...
A new hardware platform for AI accelerators capable of handling significant workloads with reduced energy requirement has been developed. The platform leverages III-V compound semiconductors to create photonic integrated circuits, which operate at the speed of light with minimal energy loss.
A new method combines ECGI with digital twins to locate the origin of premature ventricular contractions, improving accuracy by an average of 7.8 mm. The method has been applied in a real clinical case and is expected to facilitate planning interventions and reduce treatment costs.
Full Waveform Inversion (FWI) technology provides unprecedented precision in seismic imaging, breaking resolution limitations of traditional methods. It characterizes complex structures within the Earth's interior and offers higher-resolution subsurface models.
A team led by Dr. Marcus Botacin is creating a large language model (LLM) to automatically identify malware and write rules to defend against it. The LLM will use signatures to complement human analysts' skills, identifying malware faster and more accurately.
Researchers developed a nanoplatform that combines photothermal, photodynamic therapies with imaging capabilities to treat cancer. The platform achieved synergistic anticancer effects by inducing a switchable response in the tumor microenvironment.
A new hybrid microscope allows scientists to image the full 3D orientation and position of an ensemble of molecules, such as labeled proteins inside cells. This can reveal the real biology hidden from just a position change of a molecule alone.
Researchers propose H-CAP method to capture ultrafast phenomena with large frame numbers, improving spatial resolution. The method enhances high-frequency information, allowing for high-fidelity reconstruction of transient processes.
A new single-photon time-of-flight LiDAR system can acquire high-resolution 3D images of objects or scenes up to 1 kilometer away, offering improved surveillance and monitoring capabilities. The system uses a superconducting nanowire single-photon detector and achieves a higher spatial resolution than previous systems.
Researchers developed a novel microscopy technique to study metabolic changes in individual cancer cells at the single-cell level. They found that radiation treatment caused significant metabolic shifts in head and neck squamous cell carcinoma cells, particularly through the activation of HIF-1α.
A recent study published in JAMA Network Open found that heavy cannabis users exhibited reduced brain activity during working memory tasks, associated with worse performance. Abstaining from cannabis before cognitive tasks may help improve performance.
The open-source AI model analyzes medical images, generates detailed reports, and answers clinical questions to streamline diagnostics and improve accuracy. BiomedGPT aims to democratize healthcare and reduce disparities amongst patients by providing easily accessible data to bolster underserved hospitals.
A global study will explore structural changes in the brains of people with bipolar disorder and compare them to major depression. The project aims to develop new tools for diagnosis and treatment, leveraging unprecedented sample sizes from over 3,500 individuals with BD and 9,000 healthy controls.
Professors Scott Acton and Mathews Jacob of UVA's Charles L. Brown Department of Electrical and Computer Engineering were named to the IEEE Signal Processing Society's 2025 Class of Distinguished Lecturers for their groundbreaking work in signal processing, artificial intelligence, and medical imaging.
A team of researchers from the University of Ottawa has developed innovative methods to enhance frequency conversion of terahertz (THz) waves in graphene-based structures, unlocking new potential for faster, more efficient technologies in wireless communication and signal processing. These advancements hold great promise for wireless c...
The KAIST team developed a next-generation neuromorphic semiconductor-based integrated system that can learn and correct errors on its own. This technology will revolutionize AI use in everyday devices, making them faster, more private and energy-efficient.