A5039735204- Image Enhancement Techniques
- 3D Surveying and Cultural Heritage
- Remote Sensing and LiDAR Applications
- Advanced Image Fusion Techniques
- Advanced Vision and Imaging
- Advanced Image Processing Techniques
- BIM and Construction Integration
- Video Surveillance and Tracking Methods
- Visual Attention and Saliency Detection
- Conservation Techniques and Studies
- Forest ecology and management
- Knee injuries and reconstruction techniques
- 3D Modeling in Geospatial Applications
- Forest Management and Policy
- Image Processing and 3D Reconstruction
- Cultural Heritage Management and Preservation
- Image and Object Detection Techniques
- Tree Root and Stability Studies
- Remote-Sensing Image Classification
- Advanced Neural Network Applications
- Tendon Structure and Treatment
- Adversarial Robustness in Machine Learning
- Color Science and Applications
- Total Knee Arthroplasty Outcomes
- Remote Sensing and Land Use
Dongbei University of Finance and Economics
2025
Beijing University of Civil Engineering and Architecture
2024
The Fourth People's Hospital of Ningxia Hui Autonomous Region
2024
Zhejiang University
2020-2023
Dalian University of Technology
2020-2022
Luxun Academy of Fine Arts
2016
ABSTRACT Monocular 3D object detection plays a pivotal role in the field of autonomous driving and numerous deep learning‐based methods have made significant breakthroughs this area. Despite advancements accuracy efficiency, these models tend to fail when faced with adversarial attacks, rendering them ineffective. Therefore, bolstering robustness has become critical issue. To mitigate issue, we propose depth‐aware robust training method for monocular detection, dubbed DART3D. Specifically,...
Abstract Architectural heritage health assessment is the basis of scientific repair and maintenance. However, existing methods do not adequately take into account fuzziness, randomness uncertainties unique to architectural assessment. In this paper, a new evaluation model VM-NCM constructed by combining variable weight theory normal cloud theory. The enables combination qualitative ratings quantitative calculation, deals with fuzziness in process, resolves reflects uncertainty certain...
In this paper, we propose a novel sparse coding and counting method under Bayesian framework for visual tracking. contrast to existing methods, the proposed employs combination of L0 L1 norm regularize linear coefficients incrementally updated basis. The sparsity constraint enables tracker effectively handle difficult challenges, such as occlusion or image corruption. To achieve real-time processing, fast efficient numerical algorithm solving model. Although it is an NP-hard problem,...
<title>Abstract</title> Exploring the comparison and evaluation of wood structure realistic model based on fusion multi-source remote sensing data suitability reinforcement effects, aiming to proactively accurately prevent potential structural issues ensure long-term effective preservation historical buildings. Various point cloud from different sources were collected using equipment such as unmanned aerial vehicle (UAV) LiDAR; Based data, a hybrid was established, existing successfully...
Abstract This study investigates the geometric modelling of architectural heritage digital twins constructed based on multi-source point cloud data and its effectiveness in structural reinforcement assessment. Particular emphasis has been placed use static stiffness rules to identify areas weakness models need for their reinforcement, order prevent potential problems ensure long-term preservation built heritage. Taking Yingxian wooden pagoda as a case, collection data, twin model is through...
The procedure of anterior cruciate ligament (ACL) allograft preparation can be divided into fresh-frozen method (FF-allograft) or freeze-dried (FD-allograft). This study aims to biomechanically and histologically compare the graft bone tunnel integration between two allografts. In-vitro results indicated that FF-allograft FD-allograft showed excellent biocompatibility biomechanics, while a denser collagen fiber arrangement than autograft. Then, in-vivo preformation FF-allograft,...
To see clearly in low-light scenarios, a series of learning-based techniques have been developed to improve visual quality. However, due the absence semantic-level features, existing methods are perhaps less effective on semantic-oriented analysis tasks (e.g., saliency detection). break down limitation, we propose new classification-driven enhancement method with heterogeneous feature fusion. Specifically, construct image network by integrating features acquired from pre-trained...
Recently, utilizing deep networks to improve the visual quality of low-light images has attracted widespread attention. These works attach importance design network architecture and loss functions. However, they heavily overlook strong portrayal ability a critical factor that decides performance networks, i.e., training strategy. In this work, we kind effective strategy, rather than meticulously designing architectures Concretely, define lightweight illumination estimation which cannot...
In recent years, learning-based low-light image enhancement methods have shown excellent performance, but the heuristic design adopted by most requires high engineering skills for developers, causing expensive inference costs that are unfriendly to hardware platform. To handle this issue, we propose automatically discover an efficient architecture, called progressive attentive Retinex network (PAR-Net). We define a new framework introducing attention mechanism strengthen structural...
The correction of exposure-related issues is a pivotal component in enhancing the quality images, offering substantial implications for various computer vision tasks. Historically, most methodologies have predominantly utilized spatial domain recovery, limited consideration to potentialities frequency domain. Additionally, there has been lack unified perspective towards low-light enhancement, exposure correction, and multi-exposure fusion, complicating impeding optimization image processing....
In recent years, visual perception tasks (e.g., object detection) in harsh scenes low-light scenes) have received extensive attention due to their importance various practical applications autonomous driving). However, when addressing the above tasks, existing works are often too simplistic and straightforward characterize different types of features, only focus on a single task while neglecting performance multiple tasks. To break limitations methods, this paper proposes Structure...