A5036361611- Advanced SAR Imaging Techniques
- Synthetic Aperture Radar (SAR) Applications and Techniques
- Remote Sensing and Land Use
- Second Language Learning and Teaching
- Topic Modeling
- Remote-Sensing Image Classification
- Advanced Image and Video Retrieval Techniques
- Digital Communication and Language
- Advanced Sensor and Control Systems
- Microwave Imaging and Scattering Analysis
- Domain Adaptation and Few-Shot Learning
- Optical measurement and interference techniques
- Natural Language Processing Techniques
- Optical Systems and Laser Technology
- Nutrition and Health in Aging
- Cardiac Health and Mental Health
- Gender Studies in Language
- Data Quality and Management
- Impact of Technology on Adolescents
- Political theory and Gramsci
- Traditional Chinese Medicine Studies
- Mobile Learning in Education
- Media, Gender, and Advertising
- Geophysical Methods and Applications
- Anesthesia and Neurotoxicity Research
First Affiliated Hospital of Xiamen University
2024-2025
Inner Mongolia University of Technology
2022-2024
Beijing Foreign Studies University
2022-2024
Baidu (China)
2019
University College London
2018-2019
Peking University
2014-2015
Inner Mongolia University
2011-2013
Knowledge graph embedding, which projects symbolic entities and relations into continuous vector spaces, is gaining increasing attention. Previous methods allow a single static embedding for each entity or relation, ignoring their intrinsic contextual nature, i.e., may appear in different contexts, accordingly, exhibit properties. This work presents Contextualized Graph Embedding (CoKE), novel paradigm that takes account such learns dynamic, flexible, fully contextualized relation...
Attention mechanisms are widely used in Visual Question Answering (VQA) to search for visual clues related the question. Most approaches train attention models from a coarse-grained association between sentences and images, which tends fail on small objects or uncommon concepts. To address this problem, paper proposes multi-grained method. It learns explicit word-object correspondence by two types of word-level complementary sentence-image association. Evaluated VQA benchmark, model achieves...
The inherent limitation between azimuth resolution and range swath width in conventional spaceborne synthetic aperture radar (SAR) systems can be overcome by introducing the displaced phase center antenna technique. In these SAR systems, echoes from all subapertures should reconstructed combined together before single-channel processors. However, multichannel sliding spotlight Terrain Observation Progressive Scans (TOPS) modes, beam progressive sweeping during whole acquisition interval...
Abstract This study explores the representation of gender through social interactions presented in four sets English textbooks used China. By examining both stereotypes and changing roles, results reveal that although traditional exist Chinese secondary school-level textbooks, women’s roles are acknowledged they now beginning to be shown engaging non-traditional female jobs. However, family scenarios, women still depicted as taking sole responsibility for housekeeping duties without any...
This paper presents the system SSMT measuring semantic similarity between a paragraph and sentence submitted to SemEval 2014 task3: Cross-level Semantic Similarity. The special difficulty of this task is length disparity two comparison texts. We adapt several machine translation evaluation metrics for features cope with difficulty, then train regression model prediction. straightforward in intuition easy implementation. Our best run gets 0.808 Pearson correlation. METEORderived are most...
The reactions of education systems to the global lockdowns implemented during COVID-19 epidemic highlighted that there remain questions regarding how everyday technologies might be used support mass education. This paper draws on Conversation Analysis in online textual communication study key features mobile text by analysing book discussions among adult students an reading programme. We captured and analysed three patterns interaction (i.e., single linear conversations; intertwined...
Change detection is an important application of remote sensing image interpretation, which identifies changed areas interest from a pair bi-temporal images. Various deep-learning-based approaches have demonstrated promising results and most these models used encoder–decoder shape such as U-Net for segmentation areas. In order to obtain more refined features, this paper introduces change model with cascaded U-Net. The proposed network architecture contains four U-Nets ConvNeXT blocks. With...
For the urgent requirement of high resolution and wide swath, in order to observe sensitive area same time, on base combining azimuth muti-aperture with range digital beam-forming, multiple strip mode is proposed, which can attain as swath Scan mode. Through practical examples parameter design, feasibility proved.
Object detection in remote sensing images is a challenging task field of computer vision since performance negatively influenced by complicated background and various object size. However, most studies have only focused on appearance, with few taking into account scene information, which closely related to existence category objects. In this paper, we put forward new method integrating information aim generating more powerful feature. Specifically, made use GRU cell, special kind RNN, order...