A5021459831- Advanced Image Processing Techniques
- Advanced Vision and Imaging
- Image Processing Techniques and Applications
- Image and Signal Denoising Methods
- Advanced Image and Video Retrieval Techniques
- Human Pose and Action Recognition
- Anomaly Detection Techniques and Applications
- Advanced Image Fusion Techniques
- Video Surveillance and Tracking Methods
- Robotics and Sensor-Based Localization
- Interconnection Networks and Systems
- Image Enhancement Techniques
- Domain Adaptation and Few-Shot Learning
- 3D Shape Modeling and Analysis
- Stochastic Gradient Optimization Techniques
- Parallel Computing and Optimization Techniques
- Sparse and Compressive Sensing Techniques
- Advanced Data Storage Technologies
- Distributed Sensor Networks and Detection Algorithms
- Ophthalmology and Visual Impairment Studies
- Face and Expression Recognition
- Underwater Vehicles and Communication Systems
- Distributed and Parallel Computing Systems
- Advanced Neural Network Applications
- Cognitive Radio Networks and Spectrum Sensing
Sun Yat-sen University
2011-2025
Minjiang University
2019-2025
Southern Medical University Shenzhen Hospital
2025
Wuyi University
2024
Xiamen University of Technology
2018-2024
Wuhan Puai Hospital
2010-2023
Huazhong University of Science and Technology
2008-2023
Changchun University of Science and Technology
2022
Jinan University
2022
ETH Zurich
2022
This paper reviews the 2nd NTIRE challenge on single image super-resolution (restoration of rich details in a low resolution image) with focus proposed solutions and results. The had 4 tracks. Track 1 employed standard bicubic downscaling setup, while Tracks 2, 3 realistic unknown downgrading operators simulating camera acquisition pipeline. were learnable through provided pairs high train images. tracks 145, 114, 101, 113 registered participants, resp., 31 teams competed final testing...
Sparse coding has been widely applied to learning-based single image super-resolution (SR) and obtained promising performance by jointly learning effective representations for low-resolution (LR) high-resolution (HR) patch pairs. However, the resulting HR images often suffer from ringing, jaggy, blurring artifacts due strong yet ad hoc assumptions that LR representation is equal to, linear with, lies on a manifold similar or same support set as corresponding representation. Motivated success...
Recently, a number of CNN based methods have made great progress in single image super-resolution. However, these existing architectures commonly build massive network layers, bringing high computational complexity and heavy memory consumption, which is inappropriate to be applied on embedded terminals such as mobile platforms. In order solve this problem, we propose hybrid Transformer (HNCT) for lightweight general, HNCT consists four parts, are shallow feature extraction module, Hybrid...
Image super-resolution is one of the most popular computer vision problems with many important applications to mobile devices. While solutions have been proposed for this task, they are usually not optimized even common smartphone AI hardware, mention more constrained smart TV platforms that often supporting INT8 inference only. To address problem, we introduce first Mobile challenge, where target develop an end-to-end deep learning-based image can demonstrate a real-time performance on or...
In magnetic resonance imaging (MRI), the super-resolution technology has played a great role in improving image quality. The aim of this paper is to improve edges brain MRI by incorporating gradient information another contrast high-resolution image. Multi-contrast images are assumed possess same direction local pattern. We proposed establish relation model value between different restore from its input low-resolution version. similarity patches employed estimate intensity parameters,...
Example-based color transfer is a critical operation in image editing but easily suffers from some corruptive artifacts the mapping process. In this paper, we propose novel unified framework with suppression, which performs iterative probabilistic self-learning filtering scheme and multiscale detail manipulation minimizing normalized Kullback-Leibler distance. First, an applied to construct relationship between reference target images. Then, into process prevent extract details. The...
This study aims to evaluate the distribution of preoperative corneal parameters obtained using Pentacam anterior segment analyzer in Chinese male and female patients with cataracts investigate correlation between these related factors. Preoperative examination data eyes 1,255 who underwent cataract surgery were retrospectively analyzed. The AXL was used extract measurements, total measurement average age 52.9 ± 21.3 years. mean simulated keratometry values curvature refractive power...
Clinical trials are essential for discovering new treatments and advancing medical knowledge. However, the high uncertainty of carrying out clinical often ends with ineffective results. Therefore, accurate prediction trial outcomes has become a significant challenge. Numerous publicly accessible reports have been discovered to be beneficial in alleviating this challenge but lack necessary annotations formal datasets deep model training. To address issue, paper proposes construct dataset by...
Progresses has been witnessed in single image superresolution which the low-resolution images are simulated by bicubic downsampling. However, for complex degradation wild such as downsampling, blurring, noises, and geometric deformation, existing methods do not work well. Inspired a persistent memory network proven to be effective restoration, we implement core idea of human on deep residual convolutional neural network. Two types blocks designed NTIRE2018 challenge. We embed two framework...
Convolutional neural network (CNN) has achieved great success in the compressed sensing-based magnetic resonance imaging (CS-MRI). Latest deep networks for CS-MRI usually consist of a stack sub-networks, each which refines former image prediction to more accurate one. However, as sub-network number increases, information prior sub-networks little influence on subsequent ones, increases training difficulties and limits reconstruction performance model. In this paper, we propose novel network,...
Deep neural networks have made remarkable progresses on various computer vision tasks. Recent works shown that depth, width and shortcut connections of are all vital to their performances. In this paper, we introduce a method sparsify DenseNet which can reduce L-layer from O(L^2) O(L), thus simultaneously increase in more parameter-efficient computation-efficient way. Moreover, an attention module is introduced further boost our network's performance. We denote network as SparseNet. evaluate...