Search

Your search keyword '"Wang, Deming"' showing total 1,158 results
Start Over Author "Wang, Deming"
1,158 results on '"Wang, Deming"'

1. VisionUnite: A Vision-Language Foundation Model for Ophthalmology Enhanced with Clinical Knowledge

Author

Li, Zihan, Song, Diping, Yang, Zefeng, Wang, Deming, Li, Fei, Zhang, Xiulan, Kinahan, Paul E., and Qiao, Yu

Subjects
Electrical Engineering and Systems Science - Image and Video Processing, Computer Science - Artificial Intelligence, Computer Science - Computation and Language, Computer Science - Computer Vision and Pattern Recognition
Abstract

The need for improved diagnostic methods in ophthalmology is acute, especially in the less developed regions with limited access to specialists and advanced equipment. Therefore, we introduce VisionUnite, a novel vision-language foundation model for ophthalmology enhanced with clinical knowledge. VisionUnite has been pretrained on an extensive dataset comprising 1.24 million image-text pairs, and further refined using our proposed MMFundus dataset, which includes 296,379 high-quality fundus image-text pairs and 889,137 simulated doctor-patient dialogue instances. Our experiments indicate that VisionUnite outperforms existing generative foundation models such as GPT-4V and Gemini Pro. It also demonstrates diagnostic capabilities comparable to junior ophthalmologists. VisionUnite performs well in various clinical scenarios including open-ended multi-disease diagnosis, clinical explanation, and patient interaction, making it a highly versatile tool for initial ophthalmic disease screening. VisionUnite can also serve as an educational aid for junior ophthalmologists, accelerating their acquisition of knowledge regarding both common and rare ophthalmic conditions. VisionUnite represents a significant advancement in ophthalmology, with broad implications for diagnostics, medical education, and understanding of disease mechanisms.

Published
2024

4. Three-Filters-to-Normal+: Revisiting Discontinuity Discrimination in Depth-to-Normal Translation

Author

Yang, Jingwei, Xue, Bohuan, Feng, Yi, Wang, Deming, Fan, Rui, and Chen, Qijun

Subjects
Computer Science - Robotics, Computer Science - Computer Vision and Pattern Recognition
Abstract

This article introduces three-filters-to-normal+ (3F2N+), an extension of our previous work three-filters-to-normal (3F2N), with a specific focus on incorporating discontinuity discrimination capability into surface normal estimators (SNEs). 3F2N+ achieves this capability by utilizing a novel discontinuity discrimination module (DDM), which combines depth curvature minimization and correlation coefficient maximization through conditional random fields (CRFs). To evaluate the robustness of SNEs on noisy data, we create a large-scale synthetic surface normal (SSN) dataset containing 20 scenarios (ten indoor scenarios and ten outdoor scenarios with and without random Gaussian noise added to depth images). Extensive experiments demonstrate that 3F2N+ achieves greater performance than all other geometry-based surface normal estimators, with average angular errors of 7.85$^\circ$, 8.95$^\circ$, 9.25$^\circ$, and 11.98$^\circ$ on the clean-indoor, clean-outdoor, noisy-indoor, and noisy-outdoor datasets, respectively. We conduct three additional experiments to demonstrate the effectiveness of incorporating our proposed 3F2N+ into downstream robot perception tasks, including freespace detection, 6D object pose estimation, and point cloud completion. Our source code and datasets are publicly available at https://mias.group/3F2Nplus.

Published
2023

5. TransPose: 6D Object Pose Estimation with Geometry-Aware Transformer

Author

Lin, Xiao, Wang, Deming, Zhou, Guangliang, Liu, Chengju, and Chen, Qijun

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Estimating the 6D object pose is an essential task in many applications. Due to the lack of depth information, existing RGB-based methods are sensitive to occlusion and illumination changes. How to extract and utilize the geometry features in depth information is crucial to achieve accurate predictions. To this end, we propose TransPose, a novel 6D pose framework that exploits Transformer Encoder with geometry-aware module to develop better learning of point cloud feature representations. Specifically, we first uniformly sample point cloud and extract local geometry features with the designed local feature extractor base on graph convolution network. To improve robustness to occlusion, we adopt Transformer to perform the exchange of global information, making each local feature contains global information. Finally, we introduce geometry-aware module in Transformer Encoder, which to form an effective constrain for point cloud feature learning and makes the global information exchange more tightly coupled with point cloud tasks. Extensive experiments indicate the effectiveness of TransPose, our pose estimation pipeline achieves competitive results on three benchmark datasets., Comment: Accepted by NEUROCOMPUTING

Published
2023

27. Portrait Matting via Semantic and Detail Guidance

Author

Wang, Deming, Gao, Changxin, Sang, Nong, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Lu, Huimin, editor, Blumenstein, Michael, editor, Cho, Sung-Bae, editor, Liu, Cheng-Lin, editor, Yagi, Yasushi, editor, and Kamiya, Tohru, editor

Published
2023
Full Text
View/download PDF

32. High Myopia Normative Database of Peripapillary Retinal Nerve Fiber Layer Thickness to Detect Myopic Glaucoma in a Chinese Population

Author

Zhang, Xiulan, Liu, Yizhi, Lv, Lin, Friedman, David S., Jonas, Jost B., Aung, Tin, Chen, Shida, Wang, Wei, Lin, Fengbin, Song, Yunhe, Wang, Peiyuan, Li, Fei, Gao, Kai, Liu, Bingqian, Liu, Yuhong, Chen, Meiling, Bressler, Neil M., Park, Ki Ho, Lam, Dennis S.C., He, Mingguang, Ohno-Matsui, Kyoko, Weinreb, Robert N., Cheng, Ching-Yu, Healey, Paul, Zangwill, Linda M., Chen, Xiang, Tang, Guangxian, Jin, Ling, Chong, Rachel S., Ran, An Ran, Wang, Zhenyu, Jiang, Jingwen, Kong, Kangjie, Sun, Jian, Wang, Deming, Tham, Clement C., Sun, Xiaodong, and Cheung, Carol Y.

Published
2023
Full Text
View/download PDF

36. Ironic Effects of Thought Suppression: A Meta-Analysis

Author

Wang, Deming, Hagger, Martin S, and Chatzisarantis, Nikos LD

Subjects
Attention, Awareness, Defense Mechanisms, Humans, Mental Recall, Thinking, thought suppression, ironic effect, immediate enhancement effect, rebound effect, cognitive load, Psychology, Cognitive Sciences, Social Psychology
Abstract

The ironic effect of thought suppression refers to the phenomenon in which individuals trying to rid their mind of a target thought ironically experience greater levels of occurrence and accessibility of the thought compared with individuals who deliberately concentrate on the thought (Wegner, 1994, doi:10.1037/0033-295X.101.1.34). Ironic effects occurring after thought suppression, also known as rebound effects, were consistently detected by previous meta-analyses. However, ironic effects that occur during thought suppression, also known as immediate enhancement effects, were found to be largely absent. In this meta-analysis, we test Wegner's original proposition that detection of immediate enhancement effects depends on the cognitive load experienced by individuals when enacting thought suppression. Given that thought suppression is an effortful cognitive process, we propose that the introduction of additional cognitive load would compete for the allocation of existing cognitive resources and impair capacity for thought suppression. Studies (k = 31) consistent with Wegner's original thought-suppression paradigm were analyzed. Consistent with our predictions, rebound effects were observed regardless of cognitive load, whereas immediate enhancement effects were observed only in the presence of cognitive load. We discuss implications in light of ironic-process theory and suggest future thought-suppression research.

Published
2020

37. Research progress on targeted drugs for vascular malformations

Author

SHEN Yuchen, WANG Deming, and FAN Xindong

Subjects
vascular malformations, signaling pathways, mammalian target of rapamycin complex 1, mitogen-activated protein kinase kinase, rapamycin, trametinib, vascular endothelial growth factor, Medicine
Abstract

Vascular malformations, which mainly occur in the head and neck region, are a group of congenital disorders that cannot involute and dilate gradually as patients grow. Traditional therapeutic strategies for vascular malformations include laser therapy, sclerotherapy, interventional embolization, surgical resection, etc. However, for some cases with a relatively larger range of lesions, traditional therapeutic strategies might fall short of the goals. With the development of molecular genetics, gene mutations are currently recognized as the root cause of the occurrence of vascular malformations. The progression of vascular malformation lesions is further promoted by the activation of related pathways. Low-flow vascular malformations mainly involve activation of the phosphatidylinositol 3-kinase (PI3K)/protein kinase B (AKT)/mammalian target of rapamycin (mTOR) pathway, whereas high-flow vascular malformations mainly involve activation of the rat sarcoma (RAS)/rapidly accelerated fibrosarcoma (RAF)/mitogen-activated protein kinase kinase (MAPKK)/extracellular-signal regulated protein kinase (ERK) pathway. Targeted drugs against relevant gene mutations and signaling pathways have also been applied in the treatment of vascular malformations, and previous studies have shown that the mTOR inhibitor rapamycin is effective and now widely used in the treatment of low-flow vascular malformations. The PI3K inhibitor alpelisib is also promising in the treatment of venous malformations, and the MAPKK inhibitor trametinib has shown good results in the treatment of arteriovenous malformations. Therefore, traditional therapies supplemented by targeted drugs may bring new breakthroughs to the treatment of vascular malformations.

Published
2023
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results