Search

Your search keyword '"Huang, Jiancheng"' showing total 312 results
Start Over Author "Huang, Jiancheng"
312 results on '"Huang, Jiancheng"'

71. Learning Image-Adaptive Lookup Tables With Spatial Awareness for Image Harmonization

Author

Huang, Yi, Dong, Yu, Zhang, He, Huang, Jiancheng, and Chen, Shifeng

Abstract

Image harmonization aims to address inharmony between the foreground and background of composite images. Current methods mainly adopt the encoder-decoder reconstruction architecture, which restricts their application to high-resolution images due to heavy resource cost. Inspired by Photoshop, whether we can leverage some adjustment tools to apply on composite images to generate harmonized ones. This paper gives an answer “yes”, and explores a new viewpoint to address image harmonization problem by lookup tables (LUTs) rather than reconstruct full images, which saves a lot of memory. Specifically, a deep model is utilized to encode the down-sampled version of the input image as style-relevant weights to fuse multiple basic LUTs into image-adaptive ones, which are employed to transform the global color of the foreground in source high-resolution composite image to be harmonious with the background. Besides, a spatial attention module is applied to learn the spatial map for revising preliminary LUTs transformation as well as raising the local spatial awareness of the image-adaptive LUTs. Extensive experiments show that our method achieves competitive performance both qualitatively and quantitatively with a smaller model size.

Published
2024
Full Text
View/download PDF

72. DISCO: Distribution-Aware Calibration for Object Detection with Noisy Bounding Boxes

Author

Zhou, Donghao, Li, Jialin, Li, Jinpeng, Huang, Jiancheng, Nie, Qiang, Liu, Yong, Gao, Bin-Bin, Wang, Qiong, Heng, Pheng-Ann, Chen, Guangyong, Zhou, Donghao, Li, Jialin, Li, Jinpeng, Huang, Jiancheng, Nie, Qiang, Liu, Yong, Gao, Bin-Bin, Wang, Qiong, Heng, Pheng-Ann, and Chen, Guangyong

Abstract

Large-scale well-annotated datasets are of great importance for training an effective object detector. However, obtaining accurate bounding box annotations is laborious and demanding. Unfortunately, the resultant noisy bounding boxes could cause corrupt supervision signals and thus diminish detection performance. Motivated by the observation that the real ground-truth is usually situated in the aggregation region of the proposals assigned to a noisy ground-truth, we propose DIStribution-aware CalibratiOn (DISCO) to model the spatial distribution of proposals for calibrating supervision signals. In DISCO, spatial distribution modeling is performed to statistically extract the potential locations of objects. Based on the modeled distribution, three distribution-aware techniques, i.e., distribution-aware proposal augmentation (DA-Aug), distribution-aware box refinement (DA-Ref), and distribution-aware confidence estimation (DA-Est), are developed to improve classification, localization, and interpretability, respectively. Extensive experiments on large-scale noisy image datasets (i.e., Pascal VOC and MS-COCO) demonstrate that DISCO can achieve state-of-the-art detection performance, especially at high noise levels., Comment: 12 pages, 9 figures

Published
2023

74. Clearance of lipid droplets by chimeric autophagy-tethering compound ameliorates the age-related macular degeneration phenotype in mice lacking APOE

Author

Zhang, Yuelu, Huang, Jiancheng, Liang, Yu, Huang, Jiaqiu, Fu, Yuhua, Chen, Ningxie, Lu, Boxun, and Zhao, Chen

Abstract

Age-related macular degeneration (AMD) is the leading cause of irreversible blindness among the elderly, and there is currently no clinical treatment targeting the primary impairment of AMD. The earliest clinical hallmark of AMD is drusen, which are yellowish spots mainly composed of lipid droplets (LDs) accumulated under the retinal pigment epithelium (RPE). However, the potential pathogenic role of this excessive LD accumulation in AMD is yet to be determined, partially due to a lack of chemical tools to manipulate LDs specifically. Here, we employed our recently developed Lipid Droplets·AuTophagy Tethering Compounds (LD∙ATTECs) to degrade LDs and to evaluate its consequence on the AMD-like phenotypes in apoe−/− (apolipoprotein E; B6/JGpt-Apoeem1Cd82/Gpt) mouse model. apoe−/− mice fed with high-fat diet (apoe−/−-HFD) exhibited excessive LD accumulation in the retina, particularly with AMD-like phenotypes including RPE degeneration, Bruch’s membrane (BrM) thickening, drusen-like deposits, and photoreceptor dysfunction. LD·ATTEC treatment significantly cleared LDs in RPE/choroidal tissues without perturbing lipid synthesis-related proteins and rescued RPE degeneration and photoreceptor dysfunction in apoe−/−-HFD mice. This observation implied a causal relationship between LD accumulation and AMD-relevant phenotypes. Mechanically, the apoe−/−-HFD mice exhibited elevated oxidative stress and inflammatory signals, both of which were mitigated by the LD·ATTEC treatment. Collectively, this study demonstrated that LD accumulation was a trigger for the process of AMD and provided entry points for the treatment of the initial insult of AMD by degrading LDs. Abbreviations: AMD: age-related macular degeneration; APOE: apolipoprotein E; ATTECs: autophagy-tethering compounds; BODIPY: boron-dipyrromethene; BrM: Bruch’s membrane; ERG: electroretinogram; HFD: high-fat diet; LD·ATTECs: Lipid Droplets·AuTophagy Tethering Compounds; LDs: lipid droplets; OA: oleic acid; OPL: outer plexiform layer; ROS: reactive oxygen species; RPE: retinal pigment epithelium

Published
2023
Full Text
View/download PDF

89. Microarray expression profile and functional analysis of circular RNAs in choroidal neovascularization

Author

Xu Kai, Huang Jiancheng, Zhao Chen, Chen Meng, Zhang Shujie, and Zhou Rongmei

Subjects
Tube formation, genetic structures, Microarray, Angiogenesis, Vascular endothelial cell proliferation, General Medicine, Computational biology, Biology, eye diseases, General Biochemistry, Genetics and Molecular Biology, microRNA, Original Article, sense organs, KEGG, DNA microarray, Proteinaceous extracellular matrix
Abstract

Choroidal neovascularization (CNV) is a leading cause of visual loss in age-related macular degeneration (AMD). However, the molecular mechanism for CNV progression is still unclear. This study aimed to identify CNV-related circular RNAs (circRNAs), a novel class of non-coding RNAs with diverse functions. A total of 117 circRNAs were differentially expressed in the murine CNV model by microarrays. Gene ontology (GO) enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis were performed to identify the functions of selected circRNAs. The host genes of these circRNAs were predicted to be targeted to neurogenesis (ontology: biological process), proteinaceous extracellular matrix (ECM) (ontology: cellular component), and binding (ontology: molecular function). Differentially expressed circRNAs-mediated regulatory networks were enriched in ECM receptor interaction. Most of the dysregulated circRNAs could potentially bind to five different miRNAs by TargetScan and miRanda. Specifically, circ_15752 was identified in this circRNAs pool which may facilitate vascular endothelial cell proliferation, migration, and tube formation, suggesting a critical role in endothelial angiogenesis. Our work suggests that dysregulated circRNAs may be involved in CNV pathogenesis and serve as potential biomarkers for CNV.

Published
2022

90. Developing a machine learning prediction algorithm for early differentiation of urosepsis from urinary tract infection.

Author

Su, Mingkuan, Guo, Jianfeng, Chen, Hongbin, and Huang, Jiancheng

Subjects
URINARY tract infections, MACHINE learning, RECEIVER operating characteristic curves, TURNAROUND time
Abstract

Early recognition and timely intervention for urosepsis are key to reducing morbidity and mortality. Blood culture has low sensitivity, and a long turnaround time makes meeting the needs of clinical diagnosis difficult. This study aimed to use biomarkers to build a machine learning model for early prediction of urosepsis. Through retrospective analysis, we screened 157 patients with urosepsis and 417 patients with urinary tract infection. Laboratory data of the study participants were collected, including data on biomarkers, such as procalcitonin, D-dimer, and C-reactive protein. We split the data into training (80%) and validation datasets (20%) and determined the average model prediction accuracy through cross-validation. In total, 26 variables were initially screened and 18 were statistically significant. The influence of the 18 variables was sorted using three ranking methods to further determine the best combination of variables. The Gini importance ranking method was found to be suitable for variable filtering. The accuracy rates of the six machine learning models in predicting urosepsis were all higher than 80%, and the performance of the artificial neural network (ANN) was the best among all. When the ANN included the eight biomarkers with the highest influence ranking, its model had the best prediction performance, with an accuracy rate of 92.9% and an area under the receiver operating characteristic curve of 0.946. Urosepsis can be predicted using only the top eight biomarkers determined by the ranking method. This data-driven predictive model will enable clinicians to make quick and accurate diagnoses. [ABSTRACT FROM AUTHOR]

Published
2023
Full Text
View/download PDF

96. An energy-based method for uniaxially compressed rocks and its implication

Author

Luo, Yong, Huang, Jiancheng, Si, Xuefeng, Lin, Feng, and Wu, Wuxing

Abstract

To obtain the precise calculation method for the peak energy density and energy evolution properties of rocks subjected to uniaxial compression (UC) before the post-peak stage, particularly at σ ≥ 0.9σc(σdenotes stress and σcis the peak strength), extensive UC and uniaxial graded cyclical loading-unloading (GCLU) tests were performed on four rock types. In the GCLU tests, four unloading stress levels were designated when σ < 0.9σcand six unloading stress levels were designated for σ ≥ 0.9σc. The variations in the elastic energy density (ue), dissipative energy density (ud), and energy storage efficiency (C) for the four rock types under GCLU tests were analyzed. Based on the variation of uewhen σ≥ 0.9σc, a method for calculating the peak energy density was proposed. The energy evolution in rock under UC condition before the post-peak stage was examined. The relationship between C0.9(Cat σ ≥ 0.9σc) and mechanical behavior of rocks was explored, and the damage evolution of rock was analyzed in view of energy. Compared with that of the three existing methods, the accuracy of the calculation method of peak energy density proposed in this study is higher. These findings could provide a theoretical foundation for more accurately revealing the failure behavior of rock from an energy perspective.

Published
2025
Full Text
View/download PDF

100. A novel active inductor with high self-resonance frequency high Q factor and independent adjustment of inductance

Author

Zhang Zhao, Zhang Sijia, Hongyun Xie, Feihu Wang, Dongyue Jin, Wan Hezhan, Wanrong Zhang, Weicong Na, and Huang Jiancheng

Subjects
Physics, Inductance, Self resonance frequency, law, Q factor, Negative resistance, Resistor, Inductor, Topology, Varicap, Negative impedance converter, law.invention
Abstract

Based on gyrator-C principle, a novel active inductor with high self-resonance frequency, high Q factor and independent adjustment of inductance is proposed in this paper. A varactor is added in parallel with the positive transconductor to improve the tunable range of inductance. A tunable resistor between positive transconductor and negative transconductor is used to generate negative resistance, thus enhance the Q factor. Furthermore, a feedback capacitor is also employed not only to generate negative conductance for enhancing the Q factor, but also to provide negative capacitance for achieving high self- resonance frequency. The results show that the self-resonance frequency is up to 12GHz, the peak Q factor is more than 1300. The peak inductance can be adjusted significantly between 25.1nH-84.8nH with the tuning range of 108.6%, whereas the variation of the peak Q factor is only 1.6%.

Published
2019
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results