A5027048094- Speech Recognition and Synthesis
- Speech and Audio Processing
- Image and Video Quality Assessment
- Target Tracking and Data Fusion in Sensor Networks
- Advanced Image Fusion Techniques
- Remote-Sensing Image Classification
- Complex Network Analysis Techniques
- Advanced Research in Science and Engineering
- Total Knee Arthroplasty Outcomes
- Remote Sensing and LiDAR Applications
- Advanced Image and Video Retrieval Techniques
- Medical Imaging and Analysis
- Industrial Vision Systems and Defect Detection
- Radar Systems and Signal Processing
- Advanced Neural Network Applications
- Advanced SAR Imaging Techniques
- Network Traffic and Congestion Control
- Traditional Chinese Medicine Studies
- Image Enhancement Techniques
- Infrared Target Detection Methodologies
- Caching and Content Delivery
- Music and Audio Processing
- Osteoarthritis Treatment and Mechanisms
- Voice and Speech Disorders
- Advanced Clustering Algorithms Research
National University of Defense Technology
2019-2024
Shanghai Artificial Intelligence Laboratory
2022
This paper reports on the NTIRE 2022 challenge perceptual image quality assessment (IQA), held in conjunction with New Trends Image Restoration and Enhancement workshop (NTIRE) at CVPR 2022. is to address emerging of IQA by processing algorithms. The output images these algorithms have completely different characteristics from traditional distortions are included PIPAL dataset used this challenge. divided into two tracks, a full-reference track similar previous new that focuses no-reference...
Semisupervised learning (SSL), such as FixMatch, has been successfully applied to remote sensing scene classification relieve the burden of data annotation. However, in some extreme settings, only very few samples available, e.g., one ten labels per scene, can be used. When meeting this "few-shot" scenario, deep model may overfitting and prone generate confusing predictions due lack strong augmentation-based perturbations. Thus, prediction's diversity collapse, discriminability exceeds...
Community detection is a vital task in many fields, such as social networks, and financial analysis, to name few. The Louvain method, the main workhorse of community detection, popular heuristic method based on modularity. But it difficult for sequential deal with large-scale graphs. In order overcome drawback, researchers have proposed several parallel methods (Parallel Method, PLM), which suffer two challenges: (1) latency information synchronization (2) communities swap. To tackle these...
Ultrasound tongue imaging is an attractive way for speech production study as it provides effective visualization the vocal tract. Automatic classification of phonetic segments (tongue shapes) from raw ultrasound data vital further interpretation. Recently, deep learning-based approaches have been adopted in this task, which required a large-scale annotated dataset training, and not easy to be obtained practical settings. Moreover, may contain many hard examples due contamination speckle...
Ultrasound tongue imaging is widely used in clinical linguistics and phonetics. Recently, deep neural networks, especially convolutional have been the interpretation analysis of ultrasound images (UTI). Despite achieving satisfactory performance, models rely on a large amount manually labeled data, which often difficult to obtain practical settings. To address this issue, paper focuses how utilize unlabeled UTI data improve performance classification task. Specifically, we explore...
Aircraft recognition has great application value, but aircraft in remote sensing images have some problems such as low resolution, poor contrasts, sharpness, and lack of details caused by the vertical view, which make very difficult. Especially when there are many kinds differences between subtle, fine-grained is more challenging. In this paper, we propose a non-locally enhanced feature fusion network(NLFFNet) attempt to full use features from discriminative parts aircraft. First, according...
To detect highly maneuvering radar targets in low signal-to-noise ratio conditions, a hybrid long-time integration method is proposed, which combines Radon-Fourier Transform (RFT), Dynamic Programming (DP), and Binary Integration (BI), named RFT-DP-BI. A Markov model with unified range-velocity quantification formulated to describe the target's motion. Based on this model, performed. Firstly, whole time divided into multiple segments coherent performed each segment via RFT. Secondly,...
Deep learning has seen dramatic improvements in remote-sensing image scene classification. However, hard categories and examples widely exist the data sets, due to intraclass diversity interclass similarity. In this letter, we propose a novel framework address these issues. Specifically, our method first trains general model obtain confusion matrix select categories. Then sampling strategy is proposed restructure training set an expert trained focus on Finally, knowledge of distilled into...
Knee osteoarthritis (OA) is a kind of common joint disease that seriously affects mental and physical health patients. Usually, doctors make knee OA clinical diagnosis by reviewing imaging, which are easily influenced their fatigue subjective factors. Thus, computer-aided in X-rays has emerged to assist doctors. Existing methods mainly adopt two-stage approach, including localization severity prediction. However, there exists several issues: (1) accuracy existing needs be further improved,...