A5028599619- Advanced Neural Network Applications
- Advanced Image and Video Retrieval Techniques
- Synthetic Aperture Radar (SAR) Applications and Techniques
- Advanced SAR Imaging Techniques
- Robotics and Sensor-Based Localization
- Polymer composites and self-healing
- Sentiment Analysis and Opinion Mining
- Advanced Graph Neural Networks
- Complex Network Analysis Techniques
- Carbon dioxide utilization in catalysis
- biodegradable polymer synthesis and properties
- Synthesis and biological activity
- Multimodal Machine Learning Applications
- Spam and Phishing Detection
- Image Retrieval and Classification Techniques
- Marine Sponges and Natural Products
- Video Surveillance and Tracking Methods
- Advanced Text Analysis Techniques
- Text and Document Classification Technologies
Aerospace Information Research Institute
2020-2024
Chinese Academy of Sciences
2020-2024
University of Chinese Academy of Sciences
2020-2024
Ningbo Institute of Industrial Technology
2024
Civil Aviation University of China
2018
Nankai University
2018
Recently, deep-learning methods have been successfully applied to the ship detection in synthetic aperture radar (SAR) images. It is still a great challenge detect multiscale SAR ships due broad diversity of scales and strong interference inshore background. Most prevalent approaches are based on anchor mechanism that uses predefined anchors search possible regions containing objects. However, settings impact their performance as well generalization ability. Furthermore, considering sparsity...
Synthetic aperture radar (SAR) ship detection plays an important role in the field of maritime security. However, certain unique imaging properties make it challenging to extract shape features ships, such as speckle noise and strong scattering interference from irrelevant objects. These factors result inaccurate localization obvious false alarms under complex large-scale inshore scenes. To address this issue, we propose region topology network (SRT-Net), which can dynamically capture...
Emotion analysis of on-line user generated textual content is important for natural language processing and social media analytics tasks. Most previous emotion approaches focus on identifying users’ emotional states from text by classifying emotions into one the finite categories, e.g., joy, surprise, anger fear. However, there exists ambiguity characteristic analysis, since a single sentence can evoke multiple with different intensities. To address this problem, we introduce distribution...
Ship detection in synthetic aperture radar (SAR) images is a significant and challenging task. Previous works mostly rely on the manually designed anchor boxes to search for region of interests, which less flexible suffers from heavy computational load. Moreover, these detectors have limited performance complex scenes due strong interference inshore background variability object imaging characteristics. In this paper, novel ship method based scattering keypoints guided network (SKG-Net)...
Aircraft detection in synthetic aperture radar (SAR) images plays a significant role dynamic monitoring and national security. Previous methods have difficulty obtaining the desirable performance due to interference of complex scenes diversity aircraft sizes. In order solve these problems, we propose an innovative scattering feature relation network (SFR-Net) this article. First, considering that strong points SAR are usually discrete, leverage proposed point module fulfill analysis...
Developing implantable medical materials with excellent comprehensive performance has important practical applications. Cardiovascular and bile ducts are characterized by various forms of diseases high morbidity mortality. One the effective treatment modalities for such is replacement surgery. Since commercially available tubular organ sites in short supply number autologous natural grafts limited, study that can be prepared tubes great significance. This reports on an polyurethane material...
The grouping process of corner-based detectors still faces two challenges: 1) Hard-grouping. If one the paired corners is wrongly estimated, goes wrong. 2) Complex pipeline. Vanilla methods regard corner as an additional stage, complicating post-processing. To eliminate these issues, we propose a novel algorithm, termed Soft-Grouping Non-Maximum Suppression (SG-NMS), which merges with NMS into whole to simplify pipeline and lift efficiency. SG-NMS flexible match varied number detected...
Corner-guided detector enjoys potential ability to yield precise bounding boxes. However, unreliable corner pairs, generated by heuristic grouping guidance, hinder the development of this detector. In paper, we propose a novel algorithm, termed as Corner Affinity, significantly boost reliability and robustness grouping. The proposed Affinity is couple two interactional factors, namely, 1) structure affinity (SA), applying generate preliminary pairs through corresponding object's shallow...