A5004685897- 3D Shape Modeling and Analysis
- Computer Graphics and Visualization Techniques
- Advanced Numerical Analysis Techniques
- Advanced Vision and Imaging
- 3D Surveying and Cultural Heritage
- Image Processing and 3D Reconstruction
- Visual Attention and Saliency Detection
- Remote Sensing and LiDAR Applications
- Industrial Vision Systems and Defect Detection
- Advanced Image and Video Retrieval Techniques
- Digital Image Processing Techniques
- Advanced Image Fusion Techniques
- Image and Object Detection Techniques
- Generative Adversarial Networks and Image Synthesis
- Manufacturing Process and Optimization
- Aesthetic Perception and Analysis
- Image and Video Quality Assessment
- Simulation and Modeling Applications
- Anomaly Detection Techniques and Applications
- Optical measurement and interference techniques
- Image and Signal Denoising Methods
- Pregnancy and Medication Impact
- Power Systems and Technologies
- Image Retrieval and Classification Techniques
- Medical Image Segmentation Techniques
Shanghai Institute of Materia Medica
2025
Chinese Academy of Sciences
2025
Hangzhou Normal University
2020-2024
Shenyang Jianzhu University
2024
Zhejiang Sci-Tech University
2018-2023
Zhejiang University
2005-2021
Zhejiang University of Technology
2004-2018
Nanjing Medical University
2017
Jiangsu Province Hospital
2017
PRG S&Tech (South Korea)
2012
Assessing the aesthetic appeal of artworks has become an active research direction recently. However, previous works mainly focus on photographs and oil paintings, there have been few attempts in predicting aesthetics Chinese ink due to their significant differences visual features, semantic principles. Aiming at this problem, we propose a comprehensive framework, named Inkthetics, quantify paintings based deep learning. Firstly, assessment dataset is built for painting images. Secondly,...
Computational image aesthetic evaluation is a computable human perception and judgment realized by machines, which has significant impact on variety of applications such as advanced search promotional exhibition painting arts. Various approaches have been proposed in copious literature trying to solve this challenging problem. However, there few attempts reviewing works from different types visual arts, due their differences features principles. In survey, we present comprehensive listing...
Abstract Efficient semantic segmentation of large-scale point cloud scenes is a fundamental and essential task for perception or understanding the surrounding 3d environments. However, due to vast amount data, it always challenging train deep neural networks efficiently also difficult establish unified model represent different shapes effectively their variety occlusions scene objects. Taking super-patch as data representation guided by its contextual information, we propose novel multiscale...
Purpose The purpose of this paper is to propose an interactive 2D–3D garment parametric pattern-making and linkage editing scheme that integrates clothing design, simulation interaction design 3D garments 2D patterns. proposed has the potential satisfy individual needs fashion industry, such as precise fit evaluation garment, style with ease allowance constrained contour lines in design. Design/methodology/approach authors first construct a model for flat pattern corresponding body...
Cross-domain clothing retrieval is an active research topic because of its massive potential applications in fashion industry. Due to the large number garment categories or styles, and different appearances caused by camera angles, shooting conditions, messy background environments, postures dressed human body, accuracy traditional consumer-to-shop scheme always low. In this paper, based on framework deep convolution neural network, a novel cross-domain method proposed using feature fusion...
This paper presents SymmSketch—a system for creating symmetric 3D free-form shapes from 2D sketches. The reconstruction task usually separates a shape into two types of components, that is, the self-symmetric component and mutual-symmetric components. Each type can be created in an intuitive manner. Using uniform symmetry plane, user first draws sketch lines each on sketching plane. z-depth information hand-drawn input sketches calculated using their property mirror to generate construction...
Shape completion for 3-D point clouds is an important issue in the literature of computer graphics and vision. We propose end-to-end shape-preserving network through encoder-decoder architecture, which works directly on incomplete can restore their overall shapes fine-scale structures. To achieve this task, we design a novel encoder that encodes information from neighboring points different orientations scales, as well decoder outputs dense uniform complete clouds. augment object dataset...
Ridgelet transform is a new directional multi-resolution and it more suitable for describing the signals with line or super-plane singularities. Finite ridgelet discrete orthonormal version of proposed by Do Vetterli (2003). However, finite only images prime-pixels length, which limitation its application in image processing. In this paper, we improve algorithm digital implementation dyadic length proposed. This method not keeps main properties but also simplifies algorithm. We first briefly...