ADAMTS5 promotes neovascularization via autophagic degradation of PEDF in proliferative diabetic retinopathy.

Proliferative diabetic retinopathy (PDR) adversely affects visual function. Extracellular matrix proteins (ECM) contribute significantly to the development of PDR. A Disintegrin and Metalloproteinase with Thrombospondin motifs 5 (ADAMTS5) is a member of ECM proteins. ADAMTS5 participates in angiogenesis and inflammation in diverse diseases. However, the role of ADAMTS5 in PDR remains elusive. Multiplex beam array technology was used to analyze vitreous humor of PDR patients and normal people. ELISA and Western blot were used to detect the expression of ADAMTS5, PEDF and autophagy related factors. Immunofluorescence assay was used to mark the expression and localization of ADAMTS5 and PEDF. The neovascularization was detected by tube formation test. Our results revealed that ADAMTS5 expression was increased in the vitreous humor of PDR patients and oxygen-induced retinopathy (OIR) mice retinas. Inhibiting ADAMTS5 alleviated pathological angiogenesis and upregulated PEDF expression in the OIR mice. In addition, ADAMTS5 inhibited PEDF secretion in ARPE-19 cells in vitro studies, thereby inhibiting the migration of HMEC-1. Mechanically, ADAMTS5 promoted the autophagic degradation of PEDF. Collectively, inhibition of ADAMTS5 during OIR suppresses pathological angiogenesis. Our study provides a new approach for resolving pathological angiogenesis in PDR.

Propofol Regulates HIF-1 α Effect of Expression of Targeted SIRT1 Signal pathway on Kidney Renal Clear Cell Carcinoma.

The objective of this study was to investigate the effect of propofol on kidney renal clear cell carcinoma (KIRC) by regulating hypoxia-inducible factor-1α (HIF-1α) expression and silencing signal regulatory factor 1 (SIRT1) signal pathway. In this regard, human KIRC cell line RCC4 was added into 0, 5 and 10 µ G/ml propofol treatment and was divided into a control group (CG), low dose group (LG) and high dose group (HG). CCK8 was used to detect the proliferative ability of the three groups of cells, ELISA was used to detect the level of inflammatory factors in the cells, Western blot was used to detect the protein expression, qPCR was used to detect the related mRNA expression level, and Transwell method was used to detect the invasive ability of the cells in vitro. The experimental results showed that propofol decreased the proliferation and invasion ability of KIRC cells, up-regulated the expression of TGF- ß 1, IL-6, TNF- α, HIF-1 α, Fas, bax and FasL, and down-regulated the expression of SIRT1 in a dose-dependent manner. It was concluded that propofol can inhibit the SIRT1 signal pathway by up-regulating the level of HIF-1α in KIRC, which can significantly decrease the proliferation and invasion ability of KIRC cells, induce apoptosis and increase the release of intracellular inflammatory factors.

Underwater image restoration via feature priors to estimate background light and optimized transmission map.

Underwater images frequently suffer from color casts and poor contrast, due to the absorption and scattering of light in water medium. To address these two degradation issues, we propose an underwater image restoration method based on feature priors inspired by underwater scene prior. Concretely, we first develop a robust model to estimate the background light according to feature priors of flatness, hue, and brightness, which can effectively relieve color distortion. Next, we compensate the red channel of color corrected image to revise the transmission map of it. Coupled with the structure-guided filter, the coarse transmission map is refined. The refined transmission map preserves the edge information while improving the contrast. Extensive experiments on diverse degradation scenes demonstrate that our method achieves superior performance against several state-of-the-art methods.

NGF and CNTF expression and regulation mechanism by miRNA in acute paralytic strabismus.

BACKGROUND: Nerve growth factor (NGF) and ciliary neurotrophic factor (CNTF) are well-known neurotrophic factors and widely used in the clinical treatment for its promotion effect on peripheral nerve regeneration. And they were also recommended for the acute paralytic strabismus treatment. However, whether the NGF and CNTF have protective effect for the extraocular muscles of acute paralytic strabismus patients is still poorly understood. PURPOSE: In this study, we want to evaluate the biological function of NGF and CNTF on the extraocular muscle cells and reveale the regulation mechanism behind it. METHODS: Firstly, the relative expression of ngf and cntf was assessed by quantitative real-time RT-PCR. Then, the influence of NGF and CNTF on the extraocular muscle cell proliferation was determined by CCK-8. The inflammatory response in muscle cells after NGF and CNTF treatment was evaluated by ELISA and ROS detection. In addition to this, the up-stream regulation of the ngf and cntf expression was also studied. The TargetScan was used for the predication of potential miRNAs targeting with ngf and cntf 30-UTR, which is soon confirmed by luciferase activity assay. RESULTS: all the results in this research indicated that NGF and CNTF could promote the muscle cell proliferation and inhibit the inflammatory levels, then exert protective effect on the muscle cell function. RESULTS: All the results in this research indicated that NGF and CNTF could promote the muscle cell proliferation and inhibit the inflammatory levels, then exert protective effect on the muscle cell function. CONCLUSION: It was conceivable that let 7-5p was the up-stream regulator of ngf and cntf, and let 7-5p might serve as a potential molecular target for acute paralytic strabismus treatment.

Nm23-H1 is involved in the repair of ionizing radiation-induced DNA double-strand breaks in the A549 lung cancer cell line.

BACKGROUND: Although originally identified as a putative metastasis suppressor, increasing studies have confirmed a possible role for Nm23-H1 in DNA repair, through the base excision repair and nucleotide excision repair pathways. In this study, we explored whether Nm23-H1 was also involved in double-strand break repair (DSBR). METHODS AND RESULTS: We constructed a stable A549-shNm23-H1 cell line with doxycycline-regulated expression of Nm23-H1, and a A549-nNm23-H1 cell line that over expressed a nucleus-localized version of Nm23-H1. Results from both lines confirmed that Nm23-H1 participated in the repair of double-strand breaks induced by X-rays, using Comet and γ-H2AX foci assays. Subsequent studies showed that Nm23-H1 activated the phosphorylation of checkpoint-related proteins including ATM serine/threonine kinase (on S1981), tumor protein p53 (on S15), and checkpoint kinase 2 (Chk2) (on T68). We also detected interactions between Nm23-H1 and the MRE11-RAD50-NBS1 (MRN) complex, as well as Ku80. Moreover, NBS1 and Ku80 levels were comparably higher in Nm23-H1 overexpressing cells than in control cells (t = 14.462, p < 0.001 and t = 5.347, p = 0.006, respectively). As Ku80 is the keystone of the non-homologous end joining (NHEJ) pathway, we speculate that Nm23-H1 promotes DSBR through NHEJ. CONCLUSIONS: The results indicate that Nm23-H1 participates in multiple steps of DSBR.

Apurinic/apyrimidinic endonuclease 1 regulates angiogenesis in a transforming growth factor ß-dependent manner in human osteosarcoma.

Angiogenesis plays an important role in tumor growth and metastasis and has been reported to be inversely correlated with overall survival of osteosarcoma patients. It has been shown that apurinic/apyrimidinic endonuclease 1 (APE1), a dually functional protein possessing both base excision repair and redox activities, is involved in tumor angiogenesis, although these mechanisms are not fully understood. Our previous study showed that the expression of transforming growth factor ß (TGFß) was significantly reduced in APE1-deficient osteosarcoma cells. Transforming growth factor ß promotes cancer metastasis through various mechanisms including immunosuppression, angiogenesis, and invasion. In the current study, we initially revealed that APE1, TGFß, and microvessel density (MVD) have pairwise correlation in osteosarcoma tissue samples, whereas TGFß, tumor size, and MVD were inversely related to the prognosis of the cohort. We found that knocking down APE1 in osteosarcoma cells resulted in TGFß downregulation. In addition, APE1-siRNA led to suppression of angiogenesis in vitro based on HUVECs in Transwell and Matrigel tube formation assays. Reduced secretory protein level of TGFß of culture medium also resulted in decreased phosphorylation of Smad3 of HUVECs. In a mouse xenograft model, siRNA-mediated silencing of APE1 downregulated TGFß expression, tumor size, and MVD. Collectively, the current evidence indicates that APE1 regulates angiogenesis in osteosarcoma by controlling the TGFß pathway, suggesting a novel target for anti-angiogenesis therapy in human osteosarcoma.

Comparison of first-line chemotherapy based on irinotecan or other drugs to treat non-small cell lung cancer in stage IIIB/IV: a systematic review and meta-analysis.

BACKGROUND: To compare the efficacy and toxicity of irinotecan-based chemotherapy (IBC) and non-irinotecan-based chemotherapy (NIBC) as first-line treatment for stage IIIB/IV non-small cell lung cancer (NSCLC). METHODS: PubMed, EMBASE, the Cochrane Central Register of Controlled Trials (CENTRAL), abstracts from the annual meetings of ASCO and the ESMO up to 2014 were searched for randomized controlled trials (RCTs) that compared IBC with NIBC. Data on overall survival (OS) and progression-free survival (PFS) were meta-analyzed to provide hazard ratios (HRs), while data on overall response rate (ORR) and frequencies of toxicity were meta-analyzed to provide relative risk ratios (RR). RESULTS: Seven RCTs (6 RCTs from Asian population and 1 from non-Asian population) involving 1473 patients with previously untreated stage IIIB/IV NSCLC were included in the meta-analysis. IBC and NIBC were associated with similar ORR (RR: 1.08, 95%CI: 0.94 to 1.23, p=0.30), OS (HR: 0.97, 95%CI: 0.88 to 1.07, p=0.56), and PFS (HR: 1.02, 95%CI: 0.97 to 1.08, p=0.38). However, the subgroups between Asian and non-Asian patients differed significantly in OS (HR: 0.94 vs 1.87, p=0.007). There was no significant difference for hematological toxicity (RR: 0.79, 95%CI: 0.60 to 1.04, p=0.09) and significant worse for non-hematological toxicity (RR: 2.28, 95%CI: 1.60 to 3.24, p<0.001), when IBC compared to NIBC. CONCLUSIONS: As the available evidence suggests that IBC and NIBC are equivalent in terms of ORR, PFS, OS, at least in Asian patients, we recommend that IBC be considered as a first-line treatment in Asian patients with stage IIIB/IV NSCLC. However, the non-hematological toxicity of IBC must be considered.

CVANet: Cascaded visual attention network for single image super-resolution.

Deep convolutional neural networks (DCNNs) have exhibited excellent feature extraction and detail reconstruction capabilities for single image super-resolution (SISR). Nevertheless, most previous DCNN-based methods do not fully utilize the complementary strengths between feature maps, channels, and pixels. Therefore, it hinders the ability of DCNNs to represent abundant features. To tackle the aforementioned issues, we present a Cascaded Visual Attention Network for SISR called CVANet, which simulates the visual attention mechanism of the human eyes to focus on the reconstruction process of details. Specifically, we first designed a trainable feature attention module (FAM) for feature-level attention learning. Afterward, we introduce a channel attention module (CAM) to reinforce feature maps under channel-level attention learning. Meanwhile, we propose a pixel attention module (PAM) that adaptively selects representative features from the previous layers, which are utilized to generate a high-resolution image. Satisfactory, our CVANet can effectively improve the resolution of images by exploring the feature representation capabilities of different modules and the visual perception properties of the human eyes. Extensive experiments with different methods on four benchmarks demonstrate that our CVANet outperforms the state-of-the-art (SOTA) methods in subjective visual perception, PSNR, and SSIM.The code will be made available https://github.com/WilyZhao8/CVANet.

A General Spatial-Frequency Learning Framework for Multimodal Image Fusion.

multimodal image fusion involves tasks like pan-sharpening and depth super-resolution. Both tasks aim to generate high-resolution target images by fusing the complementary information from the texture-rich guidance and low-resolution target counterparts. They are inborn with reconstructing high-frequency information. Despite their inherent frequency domain connection, most existing methods only operate solely in the spatial domain and rarely explore the solutions in the frequency domain. This study addresses this limitation by proposing solutions in both the spatial and frequency domains. We devise a Spatial-Frequency Information Integration Network, abbreviated as SFINet for this purpose. The SFINet includes a core module tailored for image fusion. This module consists of three key components: a spatial-domain information branch, a frequency-domain information branch, and a dual-domain interaction. The spatial-domain information branch employs the spatial convolution-equipped invertible neural operators to integrate local information from different modalities in the spatial domain. Meanwhile, the frequency-domain information branch adopts a modality-aware deep Fourier transformation to capture the image-wide receptive field for exploring global contextual information. In addition, the dual-domain interaction facilitates information flow and the learning of complementary representations. We further present an improved version of SFINet, SFINet++, that enhances the representation of spatial information by replacing the basic convolution unit in the original spatial domain branch with the information-lossless invertible neural operator. We conduct extensive experiments to validate the effectiveness of the proposed networks and demonstrate their outstanding performance against state-of-the-art methods in two representative multimodal image fusion tasks: pan-sharpening and depth super-resolution. The source code is publicly available at https://github.com/manman1995/Awaresome-pansharpening.

Rethinking Pan-Sharpening in Closed-Loop Regularization.

It is generally known that pan-sharpening is fundamentally a PAN-guided multispectral (MS) image super-resolution problem that involves learning the nonlinear mapping from low-resolution (LR) to high-resolution (HR) MS images. Since an infinite number of HR-MS images can be downsampled to produce the same corresponding LR-MS image, learning the mapping from LR-MS to HR-MS image is typically ill-posed and the space of the possible pan-sharpening functions can be extremely large, making it difficult to estimate the optimal mapping solution. To address the above issue, we propose a closed-loop scheme that learns the two opposite mapping including the pan-sharpening and its corresponding degradation process simultaneously to regularize the solution space in a single pipeline. More specifically, an invertible neural network (INN) is introduced to perform a bidirectional closed-loop: the forward operation for LR-MS pan-sharpening and the backward operation for learning the corresponding HR-MS image degradation process. In addition, given the vital importance of high-frequency textures for the Pan-sharpened MS images, we further strengthen the INN by designing a specified multiscale high-frequency texture extraction module. Extensive experimental results demonstrate that the proposed algorithm performs favorably against state-of-the-art methods qualitatively and quantitatively with fewer parameters. Ablation studies also verify the effectiveness of the closed-loop mechanism in pan-sharpening. The source code is made publicly available at https://github.com/manman1995/pan-sharpening-Team-zhouman/.

Global-and-Local Collaborative Learning for Co-Salient Object Detection.

The goal of co-salient object detection (CoSOD) is to discover salient objects that commonly appear in a query group containing two or more relevant images. Therefore, how to effectively extract interimage correspondence is crucial for the CoSOD task. In this article, we propose a global-and-local collaborative learning (GLNet) architecture, which includes a global correspondence modeling (GCM) and a local correspondence modeling (LCM) to capture the comprehensive interimage corresponding relationship among different images from the global and local perspectives. First, we treat different images as different time slices and use 3-D convolution to integrate all intrafeatures intuitively, which can more fully extract the global group semantics. Second, we design a pairwise correlation transformation (PCT) to explore similarity correspondence between pairwise images and combine the multiple local pairwise correspondences to generate the local interimage relationship. Third, the interimage relationships of the GCM and LCM are integrated through a global-and-local correspondence aggregation (GLA) module to explore more comprehensive interimage collaboration cues. Finally, the intra and inter features are adaptively integrated by an intra-and-inter weighting fusion (AEWF) module to learn co-saliency features and predict the co-saliency map. The proposed GLNet is evaluated on three prevailing CoSOD benchmark datasets, demonstrating that our model trained on a small dataset (about 3k images) still outperforms 11 state-of-the-art competitors trained on some large datasets (about 8k-200k images).

PUGAN: Physical Model-Guided Underwater Image Enhancement Using GAN With Dual-Discriminators.

Due to the light absorption and scattering induced by the water medium, underwater images usually suffer from some degradation problems, such as low contrast, color distortion, and blurring details, which aggravate the difficulty of downstream underwater understanding tasks. Therefore, how to obtain clear and visually pleasant images has become a common concern of people, and the task of underwater image enhancement (UIE) has also emerged as the times require. Among existing UIE methods, Generative Adversarial Networks (GANs) based methods perform well in visual aesthetics, while the physical model-based methods have better scene adaptability. Inheriting the advantages of the above two types of models, we propose a physical model-guided GAN model for UIE in this paper, referred to as PUGAN. The entire network is under the GAN architecture. On the one hand, we design a Parameters Estimation subnetwork (Par-subnet) to learn the parameters for physical model inversion, and use the generated color enhancement image as auxiliary information for the Two-Stream Interaction Enhancement sub-network (TSIE-subnet). Meanwhile, we design a Degradation Quantization (DQ) module in TSIE-subnet to quantize scene degradation, thereby achieving reinforcing enhancement of key regions. On the other hand, we design the Dual-Discriminators for the style-content adversarial constraint, promoting the authenticity and visual aesthetics of the results. Extensive experiments on three benchmark datasets demonstrate that our PUGAN outperforms state-of-the-art methods in both qualitative and quantitative metrics. The code and results can be found from the link of https://rmcong.github.io/proj_PUGAN.html.

Dry eye disease among patients infected with the SARS-CoV-2 Omicron variant: a cross-sectional study.

AIM: To investigate the incidence of dry eye disease (DED) and relevant risk factors among patients infected with the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) Omicron variant. METHODS: This cross-sectional, observational analysis included 993 patients with corona virus disease 2019 (COVID-19) treated at the National Exhibition and Convention Center (Shanghai) Fangcang Shelter Hospital, from April 10 to May 26, 2022. Totally 944 uninfected control participants were recruited. All participants completed ocular surface disease index (OSDI) questionnaires, and DED symptoms were determined using OSDI scores. The demographic characteristics, length of hospital stay and in nasopharyngeal swabs were performed using questionnaires. SARS-CoV-2 Omicron variant infection was confirmed by nucleic acid-based detection in nasopharyngeal swabs using a 2019-nCoV nucleic acid detection kit. The risk factors for DED symptoms among patients with COVID-19 and control participants were determined by logistic regression analysis. RESULTS: Patients with COVID-19 showed a higher incidence of DED than controls (64.9% vs 55.1%, P<0.001). SARS-CoV-2 infection [odds ratios (ORs) (95%CI): 1.271 (1.038, 1.556)], use of contact lenses [ORs (95%CI): 9.350 (3.676, 23.783)], history of corneal refractive surgery [ORs (95%CI): 2.047 (1.494, 2.804)], poor sleep quality [ORs (95%CI): 2.657 (2.029, 3.480)], and video display terminal (VDT) use for more than 8h per day [ORs (95%CI): 6.348 (4.720, 8.538)] were found to be risk factors for DED symptoms in patients with COVID-19 as well as controls. For patients with COVID-19, the length of hospital stay [ORs (95%CI): 1.196 (1.134, 1.262)], use of contact lenses [ORs (95%CI): 20.423 (2.680, 155.632)], history of corneal refractive surgery [ORs (95%CI): 2.166 (1.321, 3.553)], poor sleep quality [ORs (95%CI): 3.650 (2.381, 5.597)], and VDT use for more than 8h per day [ORs (95%CI): 7.740 (4.918, 12.180)] were significant risk factors for DED symptoms. CONCLUSION: Patients with COVID-19 are more prone to develop symptomatic DED. SARS-CoV-2 infection and length of hospital stay are important risk factors for DED symptoms.

Caspase-8 contributes to an immuno-hot microenvironment by promoting phagocytosis via an ecto-calreticulin-dependent mechanism.

BACKGROUND: Caspase-8 (Casp8) acts as an initiator in cell apoptosis signaling. However, the role of Casp8 in tuning the tumor immune microenvironment remains controversial due to the complicated crosstalk between immune-tolerogenic apoptotic cell death and immunogenic cell death cascades. METHODS: The Cancer Genome Atlas (TCGA) and publicly accessible immune checkpoint blockade (ICB)-treated cohorts were used to investigate the clinical relevance of Casp8. A tumor-bearing mouse model was used to characterize changes in the tumor microenvironment and to explore the efficacy of ICB treatment under Casp8 knockout conditions. RESULTS: By exploring TCGA datasets, we showed that the expression level of Casp8 was associated with an immuno-hot microenvironment across various solid tumor types. Casp8 deficiency leads to decreased CD8+ T cell infiltration and resistance to anti-PD-L1 therapy in a mouse model. Mechanistically, Casp8 deficiency or pharmacological disruption results in impaired ecto-calreticulin transition in tumor cells, which in turn hampers antigen presentation in draining lymph nodes. Furthermore, radiotherapy restored sensitivity to anti-PD-L1 treatment via elevated calreticulin surface expression. CONCLUSIONS: Our data revealed a causative role of Casp8 in modulating the immunogenicity of tumor cells and responsiveness to ICB immunotherapies and proposed radiotherapy as a salvage approach to overcome Casp8 deficiency-mediated ICB resistance.

The Association Between Cataract and Cognitive Functions in Older Adults: A Longitudinal Cohort Study.

BACKGROUND: The association between cataracts and cognitive functions has been reported in several studies. However, the dynamic trajectories of cognitive changes in patients with cataracts remain unelucidated. OBJECTIVE: The aim of the study was to evaluate the dynamic trajectories of cognitive changes in patients with cataracts. METHODS: This observational cohort study recruited 1,146 patients with age-related cataracts (ARC) from the Department of Ophthalmology, Daping Hospital, from September 2020 to November 2021. The cognitive functions of the patients were assessed using a Chinese version of the Telephone Interview of Cognitive Status-40 (TICS-40) test at baseline and 6 months of follow-up. The trajectories and the associated risk factors for the longitudinal cognitive decline during the 6-month follow-up were investigated. RESULTS: Patients with severe ARC [median (IQR): 0 month, 24 (22, 25); 6 months, 23 (21,25)] had lower TICS-40 scores than those with non-severe ARC [0 month, 31 (24, 33), p < 0.001; 6 months, 31 (23,33), p < 0.001] and controls [0 month, 32 (28, 35), p < 0.001; 6 months, 32 (28, 35), p < 0.001] at both baseline and 6 months of follow-up. Age (OR: 1.311, 95% CI: 1.229 to 1.398) and cataract grade (OR: 5.569, 95% CI: 2.337 to 13.273) were found to be the risk factors of cognitive decline as indicated by a decrease in the TICS-40 scores. CONCLUSION: ARC is associated with an increased risk of longitudinal cognitive decline; however, the reversibility of such declines needs to be investigated further.

The necessity of pretreatment with 0.1% pranoprofen for femtosecond-assisted cataract surgery: A single-center, randomized controlled trial.

Purpose: To explore the effect of the variation of pupil diameter (PD) and intraocular pressure (IOP) induced by femtosecond laser treatment on the subsequent phacoemulsfication and intraocular lens implantation. And whether the application of 0.1% pranoprofen could significantly reduce the miosis and increased IOP caused by femtosecond laser treatment in femtosecond laser-assisted cataract surgery (FLACS). Methods: In this study, patients were pretreated with (trial group) or without (control group) topical 0.1% pranoprofen. The PD and IOP were measured at different time points within 30 â€‹min after the completion of the femtosecond laser treatment. Results: The comparisons of the two groups showed the PD of patients pretreated with 0.1% pranoprofen was significantly larger than that of the control only at 15 â€‹min after FLACS (P â€‹= â€‹0.046), and there was no significant difference in IOP at any time point (P â€‹> â€‹0.05). Neither the ratio of significant miosis (PD â€‹≤ â€‹5 â€‹mm) nor intraocular hypertension (IOP ≥30 â€‹mmHg) was significantly different between the control group (1.72%, 6.67%) and the trial group (1%, 4.17%) (P â€‹> â€‹0.05). Conclusions: The PD and IOP of patients undergoing FLACS showed fluctuations within a small range. The rates of significant miosis and intraocular hypertension are very low, it is safe for surgeons to complete the follow-up procedures within 30 â€‹min after femtosecond laser treatment. Pretreatment with 0.1% pranoprofen exerted a slight, albeit significant prophylactic effect preventing pupil miosis. However, it provided only a limited benefit in patients undergoing FLACS without other complications.

Learning to Enhance Low-Light Image via Zero-Reference Deep Curve Estimation.

This paper presents a novel method, Zero-Reference Deep Curve Estimation (Zero-DCE), which formulates light enhancement as a task of image-specific curve estimation with a deep network. Our method trains a lightweight deep network, DCE-Net, to estimate pixel-wise and high-order curves for dynamic range adjustment of a given image. The curve estimation is specially designed, considering pixel value range, monotonicity, and differentiability. Zero-DCE is appealing in its relaxed assumption on reference images, i.e., it does not require any paired or even unpaired data during training. This is achieved through a set of carefully formulated non-reference loss functions, which implicitly measure the enhancement quality and drive the learning of the network. Despite its simplicity, we show that it generalizes well to diverse lighting conditions. Our method is efficient as image enhancement can be achieved by an intuitive and simple nonlinear curve mapping. We further present an accelerated and light version of Zero-DCE, called Zero-DCE++, that takes advantage of a tiny network with just 10K parameters. Zero-DCE++ has a fast inference speed (1000/11 FPS on a single GPU/CPU for an image of size 1200×900×3) while keeping the enhancement performance of Zero-DCE. Extensive experiments on various benchmarks demonstrate the advantages of our method over state-of-the-art methods qualitatively and quantitatively. Furthermore, the potential benefits of our method to face detection in the dark are discussed. The source code is made publicly available at https://li-chongyi.github.io/Proj_Zero-DCE++.html.

Underwater Image Enhancement With Hyper-Laplacian Reflectance Priors.

Underwater image enhancement aims at improving the visibility and eliminating color distortions of underwater images degraded by light absorption and scattering in water. Recently, retinex variational models show remarkable capacity of enhancing images by estimating reflectance and illumination in a retinex decomposition course. However, ambiguous details and unnatural color still challenge the performance of retinex variational models on underwater image enhancement. To overcome these limitations, we propose a hyper-laplacian reflectance priors inspired retinex variational model to enhance underwater images. Specifically, the hyper-laplacian reflectance priors are established with the l1/2 -norm penalty on first-order and second-order gradients of the reflectance. Such priors exploit sparsity-promoting and complete-comprehensive reflectance that is used to enhance both salient structures and fine-scale details and recover the naturalness of authentic colors. Besides, the l2 norm is found to be suitable for accurately estimating the illumination. As a result, we turn a complex underwater image enhancement issue into simple subproblems that separately and simultaneously estimate the reflection and the illumination that are harnessed to enhance underwater images in a retinex variational model. We mathematically analyze and solve the optimal solution of each subproblem. In the optimization course, we develop an alternating minimization algorithm that is efficient on element-wise operations and independent of additional prior knowledge of underwater conditions. Extensive experiments demonstrate the superiority of the proposed method in both subjective results and objective assessments over existing methods. The code is available at: https://github.com/zhuangpeixian/HLRP.

Under-Display Camera Image Enhancement via Cascaded Curve Estimation.

The new trend of full-screen devices encourages manufacturers to position a camera behind a screen, i.e., the newly-defined Under-Display Camera (UDC). Therefore, UDC image restoration has been a new realistic single image enhancement problem. In this work, we propose a curve estimation network operating on the hue (H) and saturation (S) channels to perform adaptive enhancement for degraded images captured by UDCs. The proposed network aims to match the complicated relationship between the images captured by under-display and display-free cameras. To extract effective features, we cascade the proposed curve estimation network with sharing weights, and we introduce a spatial and channel attention module in each curve estimation network to exploit attention-aware features. In addition, we learn the curve estimation network in a semi-supervised manner to alleviate the restriction of the requirement for amounts of labeled images and improve the generalization ability for unseen degraded images in various realistic scenes. The semi-supervised network consists of a supervised branch trained on labeled data and an unsupervised branch trained on unlabeled data. To train the proposed model, we build a new dataset comprised of real-world labeled and unlabeled images. Extensive experiments demonstrate that our proposed algorithm performs favorably against state-of-the-art image enhancement methods for UDC images in terms of accuracy and speed, especially on ultra-high-definition (UHD) images.

Underwater Image Enhancement via Minimal Color Loss and Locally Adaptive Contrast Enhancement.

Underwater images typically suffer from color deviations and low visibility due to the wavelength-dependent light absorption and scattering. To deal with these degradation issues, we propose an efficient and robust underwater image enhancement method, called MLLE. Specifically, we first locally adjust the color and details of an input image according to a minimum color loss principle and a maximum attenuation map-guided fusion strategy. Afterward, we employ the integral and squared integral maps to compute the mean and variance of local image blocks, which are used to adaptively adjust the contrast of the input image. Meanwhile, a color balance strategy is introduced to balance the color differences between channel a and channel b in the CIELAB color space. Our enhanced results are characterized by vivid color, improved contrast, and enhanced details. Extensive experiments on three underwater image enhancement datasets demonstrate that our method outperforms the state-of-the-art methods. Our method is also appealing in its fast processing speed within 1s for processing an image of size 1024×1024×3 on a single CPU. Experiments further suggest that our method can effectively improve the performance of underwater image segmentation, keypoint detection, and saliency detection. The project page is available at https://li-chongyi.github.io/proj_MMLE.html.