- Image Enhancement Techniques
- Advanced Image Processing Techniques
- Advanced Image Fusion Techniques
- Infrared Target Detection Methodologies
- Advanced Vision and Imaging
- CCD and CMOS Imaging Sensors
- Remote-Sensing Image Classification
- Optical Systems and Laser Technology
- Image Processing Techniques and Applications
- Advanced Neural Network Applications
- Advanced Image and Video Retrieval Techniques
- Domain Adaptation and Few-Shot Learning
- Image and Signal Denoising Methods
- Advanced Optical Sensing Technologies
- Photoacoustic and Ultrasonic Imaging
- Artificial Intelligence in Games
- Fire Detection and Safety Systems
- Multi-Agent Systems and Negotiation
- Video Surveillance and Tracking Methods
- Photocathodes and Microchannel Plates
- Random lasers and scattering media
- Multimodal Machine Learning Applications
- Advanced Sensor and Control Systems
- Chemical Synthesis and Characterization
- Chaos control and synchronization
Technical Institute of Physics and Chemistry
2025
Samsung (China)
2023-2024
University of Leicester
2024
University of Würzburg
2023
Nanjing University of Science and Technology
2016-2022
California Maritime Academy
2021
Xi'an Jiaotong University
2004-2005
Semi-supervised semantic segmentation (SSS) has recently gained increasing research interest as it can reduce the requirement for large-scale fully-annotated training data. The current methods often suffer from confirmation bias pseudo-labelling process, which be alleviated by co-training framework. co-training-based SSS rely on hand-crafted perturbations to prevent different sub-nets collapsing into each other, but these artificial cannot lead optimal solution. In this work, we propose a...
This paper reviews the NTIRE 2023 challenge on image denoising (σ = 50) with a focus proposed solutions and results. The aim is to obtain network design capable produce high-quality results best performance measured by PSNR for denoising. Independent additive white Gaussian noise (AWGN) assumed level 50. had 225 registered participants, 16 teams made valid submissions. They gauge state-of-the-art
Limited by the number of available photons collected a pixel and stability imaging system, images captured under poor illumination conditions are often degraded heavy shot noise low contrast. In this letter, we propose novel low-light image enhancement algorithm using Poisson noise-aware Retinex model, in which distribution is considered to formulate fidelity term model for first time. Furthermore, space-variant weight maps total variation regularization terms calculated according piecewise...
We present the new Bokeh Effect Transformation Dataset (BETD), and review proposed solutions for this novel task at NTIRE 2023 Challenge. Recent advancements of mobile photography aim to reach visual quality full-frame cameras. Now, a goal in computational is optimize effect itself, which aesthetic blur out-of-focus areas an image. Photographers create by benefiting from lens optical properties.The work design neural network capable converting one another without harming sharp foreground...
In the process of removing nighttime flare, it is crucial to have a large receptive field due fact that flare can occupy substantial portion an image, even potentially entire image. However, conventional window-based Transformer approaches restrict within window, limiting its ability capture global features. And cause dark regions become brighter and result in loss contrast alteration frequency characteristics To address these challenges, we introduce FF-Former, which based on Fast Fourier...
Abstract Biomimetic nanochannels enable fast and selective transport of mononuclear metal ions; however, their construction for complex ion remains in its infancy due to the nonuniform charge distribution large geometric dimensions coordination compounds. Herein, an ionic electrostatic interaction template strategy is proposed prepare biomimetic channels capture ions. Using [UO 2 (CO 3 ) ] 4− as a template, various quaternary ammonium monomers with lengths 6.2−8.4 Å are decorated on 1D...
For the detection of weak signals, photon counting technology has attracted much attention. In this paper, a method based on is experimentally demonstrated to realize corona discharge at solar-blind ultraviolet (UV) wavelength. our setup, UV optical filter and Te-Cs cathode are used eliminate effects light with undesired wavelengths, i.e., outside band. A two-stage microchannel plate high voltage can achieve gain up 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML"...
Developing and integrating advanced image sensors with novel algorithms in camera systems are prevalent the increasing demand for computational photography imaging on mobile platforms. However, lack of high-quality data research rare opportunity in-depth exchange views from industry academia constrain development intelligent (MIPI). With success 1st MIPI Workshop@ECCV 2022, we introduce second challenge including four tracks focusing algorithms. In this paper, summarize review Nighttime...
Aiming to solve the problem of low-light-level (LLL) images with dim overall brightness, uneven gray distribution, and low contrast, in this paper, we propose an effective LLL image enhancement method based on guided filter multi-scale fusion for contrast detail preservation. First, a base image(s) are obtained by using filter. After procedure, is processed maximum entropy-based Gamma correction stretch level distribution. Unlike existing methods, enhance kernel, which reflects area...
Low-light images always suffer from dim overall brightness, low contrast, and dynamic ranges, thus result in image degradation. In this paper, we propose an effective method for low-light enhancement based on the just-noticeable-difference (JND) optimal contrast-tone mapping (OCTM) models. First, guided filter decomposes original into base detail images. After filtering, are processed visual masking model to enhance details effectively. At same time, brightness of is adjusted JND OCTM...
Bokeh effect transformation(BET) aims to transform the bokeh of one lens another without harming sharp foreground regions in image. Recent studies have shown remarkable success rendering. However, unlike traditional rendering task, BET task needs image into specified lens. The existing method is invalid or inefficient for BET, because each pair independently build different model. To address this limitation, we propose NAFBET, a scalable approach than can perform multiple using only single...
We present SemiCRL, a novel framework for volumetric medical image segmentation that formulates an innovative contrastive learning methodology in semi-supervised setting. leverage the pseudo-labels generated to guide selection of negative samples our learning, aiming alleviate class collision issue and learn enhanced class-discriminative latent representations. However, address inaccuracies pseudo-labels, which stem from empirical distribution misalignment between labeled unlabeled data, we...
Tone mapping operators (TMOs) aim to adjust high dynamic range (HDR) images low (LDR) ones so that they can be displayed on conventional devices with visual information retained. Nonetheless, existing TMOs successfully tone-map only limited types of HDR images, and the parameters need manually adjusted yield best subjective-quality tone-mapped outputs. To cope aforementioned issues, an adaptive parameter-free scene-adaptive TMO for adjusting detail enhancing is proposed a high-resolution...
The generalization of decision-making agents encompasses two fundamental elements: learning from past experiences and reasoning in novel contexts. However, the predominant emphasis most interactive environments is on learning, often at expense complexity reasoning. In this paper, we introduce CivRealm, an environment inspired by Civilization game. Civilization's profound alignment with human history society necessitates sophisticated while its ever-changing situations demand strong to...
In nonlinear topological systems, edge solitons emerge either as bifurcations of linear states or nonlinearity-induced localized without protection. Although electrical circuits have proven to be a versatile platform for realizing various types insulators, observing and their phase transition in circuit lattices remains challenging. Here, we realize quench dynamics experimentally demonstrate the existence both nontopological trimer lattice, well between these two solitons. Under weak...