A5111208115- Natural Language Processing Techniques
- Text Readability and Simplification
- Remote-Sensing Image Classification
- Robotics and Automated Systems
- Semiconductor materials and devices
- Space Technology and Applications
- Medical Image Segmentation Techniques
- Automated Road and Building Extraction
- Radiation Effects in Electronics
- Topic Modeling
- Remote Sensing and LiDAR Applications
- Currency Recognition and Detection
- Computer Graphics and Visualization Techniques
- VLSI and Analog Circuit Testing
- Advanced Vision and Imaging
University of Waterloo
2025
Tencent (China)
2024
Beijing University of Posts and Telecommunications
2023
TianQin is a future space-based gravitational wave observatory targeting the frequency window of $10^{-4}$ Hz $\sim 1$ Hz. A large variety sources are expected in this band, including merger massive black hole binaries, inspiral extreme/intermediate mass ratio systems, stellar-mass Galactic compact and so on. will consist three Earth orbiting satellites on nearly identical orbits with orbital radii about $10^5$ km. The form normal triangle constellation whose plane perpendicular to ecliptic...
Road network data are crucial for various applications, such as road planning, traffic control, map navigation, autonomous driving, and smart city construction. Automated extraction from high-spatial-resolution remotely sensed imagery has shown promise in In recent years, the advent of deep learning algorithms pushed towards auto - mation, achieving very high accuracy. However, latest models often less applied field lack comparative experiments guidance. Therefore, this research selected...
There has been much interest in accurate cryptocurrency price forecast models by investors and researchers. Deep Learning are prominent machine learning techniques that have transformed various fields shown potential for finance economics. Although deep explored forecasting, it is not clear which suitable due to high market volatility. In this study, we review the literature about forecasting evaluate novel stock prediction. Our include variants of long short-term memory (LSTM) recurrent...
In this abstract we describe a production implementation of 3D anisotropic VTI RTM using finite-difference scheme on cluster. Our end-to-end code is about four times faster than blade that consists dual Intel quadcore Harpertown running similar code, including vectorization. It interesting to note in the final disk IO bottleneck. This means compute nodes (increased FLOPS per core) will not improve performance. What needed more bandwidth (>200MB/s processor) make it run faster, and/or...
SECDED code is the combinational logic for communication of data between microprocessors and memories which commonly used in semiconductor systems due to its advantage implementation. With advances integrated circuit technology, soft error rate importance selecting optimal codes increase. This paper presents their major expansion theoretical application gives suggestions optimization needs.