A5090866370- Advanced Image and Video Retrieval Techniques
- Image Retrieval and Classification Techniques
- Domain Adaptation and Few-Shot Learning
- Visual Attention and Saliency Detection
- Video Analysis and Summarization
- Advanced Image Fusion Techniques
- Image and Video Quality Assessment
- Advanced Graph Neural Networks
- Advanced Neural Network Applications
- Multimodal Machine Learning Applications
- Video Surveillance and Tracking Methods
- Robotics and Sensor-Based Localization
- Text and Document Classification Technologies
- Industrial Vision Systems and Defect Detection
- Complex Network Analysis Techniques
- Image Enhancement Techniques
- Brain Tumor Detection and Classification
- Advanced Image Processing Techniques
- Currency Recognition and Detection
- Fire Detection and Safety Systems
- COVID-19 diagnosis using AI
- Machine Learning and ELM
- Biometric Identification and Security
- Image and Video Stabilization
- Music and Audio Processing
Xi'an University of Technology
2014-2025
University of Technology
2024
China Academy of Printing Technology
2018-2024
Xi'an Jiaotong University
2010-2013
Objective image quality assessment employs mathematical and computational theory to objectively assess the of output images based on human visual system (HVS). In this paper, a novel approach multifeature extraction in spatial frequency domains is proposed. We combine gradient magnitude phase congruency maps generate local structure (LS) map, which can perceive structural distortions. The LS matches well with HVS highlights differences details. For complex information, such as texture...
Due to the explosive increase in online videos, near-duplicate video retrieval (NDVR) has attracted much researcher attention. NDVR very wide applications, such as copyright protection, monitoring, and automatic tagging. Local features serve elementary building blocks many algorithms, most of them exploit local volume information using a bag (BOF) representation. However, representation ignores potentially valuable about global distribution interest points. Moreover, discriminative power...
Suffering from the multi-view data diversity and complexity for semi-supervised classification, most of existing graph convolutional networks focus on architecture construction or salient structure preservation, ignore complete classification contribution. To mine more distribution with consideration specificity commonality, we propose fusion based (SF-GCN) improving performance classification. SF-GCN can not only retain special characteristic each view by spectral embedding, but also...
Feature generating networks face to the most important question, which is fitting difference (inconsistence) of distribution between generated feature and real data. This inconsistence further influence performance model, because training samples from seen classes disjointed with testing unseen in zero-shot learning (ZSL). In generalization (GZSL), come not only but also for closer practical situation. Therefore, difficultly obtain satisfactory challenging GZSL by adversarial semantic...
The basis of image processing is to evaluate and monitor quality using algorithms rather than subjective methods. Conventional gradient operators have been popularly used in previous assessment tasks reflect the edge contour an image, while there are some obvious defects terms selection mask scale direction. Some improved versions also less ideal since they fail consider information same pixel different directions at time. authors adopt a powerful operator that can simultaneously capture all...
Using clustering method to detect useful patterns in large datasets has attracted considerable interest recently. The HKM algorithm (Hierarchical K-means) is very efficient large-scale data analysis. It been widely used build visual vocabulary for scale video/image retrieval system. However, the speed and even accuracy of hierarchical K-means still have room be improved. In this paper, we propose a Parallel N-path quantification which improves on following ways. Firstly, replace Euclidean...
The query image is usually a simple and single object in retrieval, the reference images database have many distractions. precision of retrieval can be greatly improved If target regions are extracted during retrieval. So this paper proposes Bow method based on SSD detection. First, training gallery manually annotated to record location size information. Second, detection model trained with labeled obtain model. Third, used locate similar graph. Finally, region information mapped into...