A5061089304- Image and Signal Denoising Methods
- Advanced Image Fusion Techniques
- Advanced Image Processing Techniques
- Video Surveillance and Tracking Methods
- Topic Modeling
- Face recognition and analysis
- Impact of Light on Environment and Health
- Text and Document Classification Technologies
- Educational Technology and Pedagogy
- Face and Expression Recognition
- Brain Tumor Detection and Classification
- Sentiment Analysis and Opinion Mining
- Photoacoustic and Ultrasonic Imaging
- Medical Image Segmentation Techniques
- Biometric Identification and Security
- Educational Technology and Assessment
- Infrared Target Detection Methodologies
Yunnan Normal University
2021-2024
Xi'an University of Architecture and Technology
2023
Abstract Image denoising is one of the hottest topics in image restoration area, it has achieved great progress both terms quantity and quality recent years, especially after wide intensive application deep neural networks. In many learning based models, performance can greatly benefit from prepared clean/noisy pairs used for model training, however, also limits these models real scenes. Therefore, more researchers tend to develop that be learned without pairs, namely well generalised...
In image denoising, deep convolutional neural networks (CNNs) can obtain favorable performance on removing spatially invariant noise. However, many of these cannot perform well the real noise (i.e. variant noise) generated during acquisition or transmission, which severely sets back their application in practical denoising tasks. Instead continuously increasing network depth, researchers have revealed that expanding width also be a useful way to improve model performance. It has been...
Aspect-Based Sentiment Analysis (ABSA), also called Aspect Level Classification (ALSC), is a common task in Natural Language Processing (NLP). mainly aims to extract and classify the sentiments objects texts. In this paper, we propose novel BERT-based ABSA model, which combines an adversarial training procedure with relational graph attention neural network (R-GAT). To our best knowledge, it first model that simultaneously using training, BERT for aspect-based sentiment analysis. proposed...
Deep convolutional neural networks (CNNs) for image denoising can effectively exploit rich hierarchical features and have achieved great success. However, many deep CNN-based models equally utilize the of noisy images without paying attention to more important useful features, leading relatively low performance. To address issue, we design a new Two-stage Progressive Residual Dense Attention Network (TSP-RDANet) denoising, which divides whole process into two sub-tasks remove noise...
Because of the rise convolutional neural networks, face detection has received a lot attention and research. Especially affected by image quality, resolution size, illumination, The small faces been major challenge for tasks. Different from general detectors, we propose an anchor- free method based on object framework YOLOX. YOLOX network incorporating mechanism is proposed problem wrong omission in at low resolution. were evaluated some open benchmark datasets, promising results obtained.
At present, transformer-based target tracking algorithms mainly use transformers to fuse deep convolution features, their accuracy is competitive, however compared with convolutional neural networks, speed slow. Due the long-distance dependence characteristics, it difficult obtain rich local information when extracting visual results may become worse, or even fail in later procedures. The partial algorithm based on Siamese network has great advantages information, its cannot fully reach...
Abstract This paper combines the text similarity calculation of co-occurring words, semantic similarity, cosine and based on a language network joins feature fusion to construct multi-feature method, so as realize resource distribution interactive teaching mode. Through design functional modules, database design, framework system, JavaEE technology is used system complete online offline teaching. After testing, resources under this algorithm does not change number users increases, range...