Yan Liang

Bio: Yan Liang is an academic researcher from Northwestern Polytechnical University. The author has contributed to research in topics: Filter (signal processing) & Gaussian. The author has an hindex of 36, co-authored 262 publications receiving 4553 citations. Previous affiliations of Yan Liang include Chinese Ministry of Education & Chinese Academy of Sciences.

Semi-coupled dictionary learning with applications to image super-resolution and photo-sketch synthesis

Abstract: In various computer vision applications, often we need to convert an image in one style into another style for better visualization, interpretation and recognition; for examples, up-convert a low resolution image to a high resolution one, and convert a face sketch into a photo for matching, etc. A semi-coupled dictionary learning (SCDL) model is proposed in this paper to solve such cross-style image synthesis problems. Under SCDL, a pair of dictionaries and a mapping function will be simultaneously learned. The dictionary pair can well characterize the structural domains of the two styles of images, while the mapping function can reveal the intrinsic relationship between the two styles' domains. In SCDL, the two dictionaries will not be fully coupled, and hence much flexibility can be given to the mapping function for an accurate conversion across styles. Moreover, clustering and image nonlocal redundancy are introduced to enhance the robustness of SCDL. The proposed SCDL model is applied to image super-resolution and photo-sketch synthesis, and the experimental results validated its generality and effectiveness in cross-style image synthesis.

Neutrophil to lymphocyte ratio (NLR) and platelet to lymphocyte ratio (PLR) were useful markers in assessment of inflammatory response and disease activity in SLE patients.

Abstract: Objective: Although there have been extensive investigations on neutrophil to lymphocyte ratio (NLR), platelet to lymphocyte ratio (PLR) and mean platelet volume (MPV) in many diseases, their roles in systemic lupus erythematosus (SLE) remain unclear. The purpose of the present study was to evaluate NLR, PLR, and MPV levels in adult SLE patients and explore their clinical significance.Methods: A retrospective study involving 154 adult SLE patients and 151 healthy controls was performed. All clinical characteristics of the SLE patients were extracted from their medical records. NLR, PLR, and MPV levels between SLE patients and healthy controls were compared, and correlations between these indexes and clinical characteristics were analyzed.Results: Increased NLR, PLR, and MPV were observed in SLE patients. NLR was positively correlated with C-reaction protein (r = 0.509, p < 0.01), erythrocyte sedimentation rate (r = 0.610, p < 0.01), and SLE Disease Activity Index (SLEDAI) scores (r = 0.471, p < 0....

SIRT1 Promotes Proliferation and Prevents Senescence Through Targeting LKB1 in Primary Porcine Aortic Endothelial Cells

Abstract: Rationale: Endothelial senescence causes endothelial dysfunction, promotes atherogenesis and contributes to age-related vascular disorders. SIRT1 is a conserved NAD+-dependent deacetylase possessing beneficial effects against aging-related diseases, despite that the detailed functional mechanisms are largely uncharacterized. Objective: The present study is designed to evaluate the protective effects of SIRT1 on endothelial senescence and to elucidate the underlying mechanisms. Methods and Results: An in vitro senescence model was established by prolonged culture of primary endothelial cells isolated from porcine aorta. The freshly isolated “young” cells gradually underwent senescence during 1 month of repetitive passages. Both mRNA and protein expressions of SIRT1 were progressively decreased. In contrast, the protein levels of LKB1, a serine/threonine kinase and tumor suppressor, and the phosphorylation of its downstream target AMPK(Thr172) were dramatically increased in senescent cells. Overexpression o...

LncRNA CASC9 promotes esophageal squamous cell carcinoma metastasis through upregulating LAMC2 expression by interacting with the CREB-binding protein.

Abstract: Esophageal squamous cell carcinoma (ESCC) is the main subtype of esophageal cancer. Long noncoding RNAs (lncRNAs) are thought to play a critical role in cancer development. Recently, lncRNA CASC9 was shown to be dysregulated in many cancer types, but the mechanisms whereby this occurs remain largely unknown. In this study, we found that CASC9 was significantly upregulated in ESCC tissues, with further analysis revealing that elevated CASC9 expression was associated with ESCC prognosis and metastasis. Furthermore, we found that CASC9 knockdown significantly repressed ESCC migration and invasion in vitro and metastasis in nude mice in vivo. A microarray analysis and mechanical experiments indicated that CASC9 preferentially affected gene expression linked to ECM–integrin interactions, including LAMC2, an upstream inducer of the integrin pathway. We demonstrated that LAMC2 was consistently upregulated in ESCC and promoted ESCC metastasis. LAMC2 overexpression partially compromised the decrease of cell migration and invasion capacity in CASC9 knockdowns. In addition, we found that both CASC9 and LAMC2 depletion reduced the phosphorylation of FAK, PI3K, and Akt, which are downstream effectors of the integrin pathway. Moreover, the reduction in phosphorylation caused by CASC9 depletion was rescued by LAMC2 overexpression, further confirming that CASC9 exerts a pro-metastatic role through LAMC2. Mechanistically, RNA pull-down and RNA-binding protein immunoprecipitation (RIP) assay indicated that CASC9 could bind with the transcriptional coactivator CREB-binding protein (CBP) in the nucleus. Chromatin immunoprecipitation (ChIP) assay additionally illustrated that CASC9 increased the enrichment of CBP and H3K27 acetylation in the LAMC2 promoter, thereby upregulating LAMC2 expression. In conclusion, we demonstrate that CASC9 upregulates LAMC2 expression by binding with CBP and modifying histone acetylation. Our research reveals the prognostic and pro-metastatic roles for CASC9 in ESCC, suggesting that CASC9 could serve as a biomarker for prognosis and a target for metastasis treatment.

Body mass index and the risk of rheumatoid arthritis: a systematic review and dose-response meta-analysis

Abstract: The evidence from published studies on the association between obesity and rheumatoid arthritis has been contradictory. To clarify the association between obesity and rheumatoid arthritis, we conducted a systematic review and dose-response meta-analysis to assess the relationship between body mass index and rheumatoid arthritis risk. A systematic literature search of PubMed and Embase (up to 12 July 2014) was performed to identify all eligible published reports. The pooled relative risk results with corresponding 95% confidence intervals of rheumatoid arthritis development were estimated using a random-effects model. Eleven eligible related citations fulfilled the inclusion criteria and were included in the study. Compared with individuals with a body mass index under 30, obese individuals showed an association with a significantly increased risk of rheumatoid arthritis (relative risk = 1.25, 95% confidence interval: 1.07 to 1.45, P heterogeneity <0.01, I2 = 63%). Compared to normal weight subjects, the pooled relative risks for rheumatoid arthritis were 1.31 (1.12 to 1.53) and 1.15 (1.03 to 1.29) for the categories of obese and overweight, respectively. In the dose-response analysis, there was evidence of a nonlinear association (P nonlinear = 0.005) and the estimated summary relative risk for a 5-unit increment was 1.03 (95% confidence interval: 1.01 to 1.05, P heterogeneity = 0.001, I2 = 70.0%). An increase in body mass index can contribute to a higher risk for rheumatoid arthritis development. However, the finding also highlights the need for research on the association between body mass index and rheumatoid arthritis risk with adjustment for more confounding factors.

Pattern Recognition and Machine Learning

Abstract: Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network

Abstract: Recently, several models based on deep neural networks have achieved great success in terms of both reconstruction accuracy and computational performance for single image super-resolution. In these methods, the low resolution (LR) input image is upscaled to the high resolution (HR) space using a single filter, commonly bicubic interpolation, before reconstruction. This means that the super-resolution (SR) operation is performed in HR space. We demonstrate that this is sub-optimal and adds computational complexity. In this paper, we present the first convolutional neural network (CNN) capable of real-time SR of 1080p videos on a single K2 GPU. To achieve this, we propose a novel CNN architecture where the feature maps are extracted in the LR space. In addition, we introduce an efficient sub-pixel convolution layer which learns an array of upscaling filters to upscale the final LR feature maps into the HR output. By doing so, we effectively replace the handcrafted bicubic filter in the SR pipeline with more complex upscaling filters specifically trained for each feature map, whilst also reducing the computational complexity of the overall SR operation. We evaluate the proposed approach using images and videos from publicly available datasets and show that it performs significantly better (+0.15dB on Images and +0.39dB on Videos) and is an order of magnitude faster than previous CNN-based methods.

Weighted Nuclear Norm Minimization with Application to Image Denoising

Abstract: As a convex relaxation of the low rank matrix factorization problem, the nuclear norm minimization has been attracting significant research interest in recent years. The standard nuclear norm minimization regularizes each singular value equally to pursue the convexity of the objective function. However, this greatly restricts its capability and flexibility in dealing with many practical problems (e.g., denoising), where the singular values have clear physical meanings and should be treated differently. In this paper we study the weighted nuclear norm minimization (WNNM) problem, where the singular values are assigned different weights. The solutions of the WNNM problem are analyzed under different weighting conditions. We then apply the proposed WNNM algorithm to image denoising by exploiting the image nonlocal self-similarity. Experimental results clearly show that the proposed WNNM algorithm outperforms many state-of-the-art denoising algorithms such as BM3D in terms of both quantitative measure and visual perception quality.

Coupled Generative Adversarial Networks

Abstract: We propose the coupled generative adversarial nets (CoGAN) framework for generating pairs of corresponding images in two different domains. The framework consists of a pair of generative adversarial nets, each responsible for generating images in one domain. We show that by enforcing a simple weight-sharing constraint, the CoGAN learns to generate pairs of corresponding images without existence of any pairs of corresponding images in the two domains in the training set. In other words, the CoGAN learns a joint distribution of images in the two domains from images drawn separately from the marginal distributions of the individual domains. This is in contrast to the existing multi-modal generative models, which require corresponding images for training. We apply the CoGAN to several pair image generation tasks. For each task, the CoGAN learns to generate convincing pairs of corresponding images. We further demonstrate the applications of the CoGAN framework for the domain adaptation and cross-domain image generation tasks.