A5101699056- Crystallization and Solubility Studies
- X-ray Diffraction in Crystallography
- Advanced Neural Network Applications
- Advanced Numerical Analysis Techniques
- Advanced Image Processing Techniques
- Adversarial Robustness in Machine Learning
- Domain Adaptation and Few-Shot Learning
- Polynomial and algebraic computation
- Infrared Target Detection Methodologies
- Digital Media Forensic Detection
- Medical Image Segmentation Techniques
- Machine Learning in Healthcare
- Software Testing and Debugging Techniques
- Advanced Mathematical Theories and Applications
- Image and Signal Denoising Methods
- Artificial Intelligence in Healthcare and Education
- Robotic Mechanisms and Dynamics
- Advanced Optimization Algorithms Research
- Software System Performance and Reliability
- Video Surveillance and Tracking Methods
- Medical Imaging and Analysis
- Software Reliability and Analysis Research
- COVID-19 diagnosis using AI
- Matrix Theory and Algorithms
- Advanced Vision and Imaging
University of Science and Technology Beijing
2024
Beijing University of Posts and Telecommunications
2023
Hefei University of Technology
2021
Microsoft (United States)
2020
Eskişehir Osmangazi University
2019
Microsoft Research Asia (China)
2019
Massachusetts Institute of Technology
2016
The Visual Object Tracking challenge VOT2019 is the seventh annual tracker benchmarking activity organized by VOT initiative. Results of 81 trackers are presented; many state-of-the-art published at major computer vision conferences or in journals recent years. evaluation included standard and other popular methodologies for short-term tracking analysis as well methodology long-term analysis. was composed five challenges focusing on different domains: (i) VOTST2019 focused RGB, (ii)...
One-shot weight sharing methods have recently drawn great attention in neural architecture search due to high efficiency and competitive performance. However, across models has an inherent deficiency, i.e., insufficient training of subnetworks hypernetworks. To alleviate this problem, we present a simple yet effective distillation method. The central idea is that can learn collaboratively teach each other throughout the process, aiming boost convergence individual models. We introduce...
We adapt the theory of normal and special polynomials from symbolic integration to summation setting, then built up a general framework embracing both usual shift case q-shift case. In context this framework, we develop unified reduction algorithm, subsequently creative telescoping applicable hypergeometric terms their q-analogues. Our algorithms allow split only when it is really necessary, thus instantly reveal intrinsic differences between these two cases. Computational experiments are...
Deep learning has achieved remarkable results in the areas of computer vision, speech recognition, natural language processing and most recently, even playing Go. The application deep-learning to problems healthcare, however, gained attention only recent years, it's ultimate place at bedside remains a topic skeptical discussion. While there is growing academic interest Machine Learning (ML) techniques clinical problems, many community see little incentive upgrade from simpler methods, such...
The analysis of corrosion images is crucial in materials science, where acquiring clear fundamental for subsequent analysis. goal deblurring metal to reconstruct from degraded ones. To the best our knowledge, this study introduces first paired blurry-sharp image dataset specifically designed domain, filling a critical gap existing research. This innovative approach effectively addresses unique challenges associated with images. We propose novel network (MCIDN) that employs dual-domain...
We present two evaluation-based algorithms: one for computing logarithmic parts and the other determining complete in transcendental function integration. Empirical results illustrate that new algorithms are markedly faster than those based respectively on resultants, contraction of ideals, subresultants Gröbner bases. They may be used to accelerate Risch's algorithm integrands, help us compute elementary integrals over towers efficiently.
Medical image registration is a critical task that estimates the spatial correspondence between pairs of images. However, current traditional and deep-learning-based methods rely on similarity measures to generate deforming field, which often results in disproportionate volume changes dissimilar regions, especially tumor regions. These can significantly alter size underlying anatomy, limits practical use clinical diagnosis. To address this issue, we have formulated with tumors as constraint...
Fault root cause localization is a crucial stage in software debugging. Spectrum-based fault methods are hot topics automated debugging research, but their effectiveness heavily relies on the quality of test cases. Most existing cases poor quality, as different target with multiple faults exhibit same crash output. This makes it difficult to establish one-to-one mapping between inputs and outputs. Consequently, application scenarios such techniques limited. In this paper, we propose case...