A5073154896- Advanced Graph Neural Networks
- Complex Network Analysis Techniques
- Brain Tumor Detection and Classification
- 3D Surveying and Cultural Heritage
- Synthetic Aperture Radar (SAR) Applications and Techniques
- Advanced Computing and Algorithms
- Quantum Chromodynamics and Particle Interactions
- Infrastructure Maintenance and Monitoring
- Advanced Neural Network Applications
- Human Pose and Action Recognition
- Neural Networks and Applications
- Recycling and utilization of industrial and municipal waste in materials production
- Bauxite Residue and Utilization
- Particle physics theoretical and experimental studies
- Recommender Systems and Techniques
- Landslides and related hazards
- Natural Language Processing Techniques
- Phase Change Materials Research
- Multimodal Machine Learning Applications
- Video Analysis and Summarization
- Computational and Text Analysis Methods
- High-Energy Particle Collisions Research
- Extraction and Separation Processes
- Topic Modeling
Shenzhen University
2025
Fuzhou University
2023-2024
University of Washington
2024
South China University of Technology
2024
Zhuhai Institute of Advanced Technology
2024
Institute of High Energy Physics
2008
The training of controllable text-to-video (T2V) models relies heavily on the alignment between videos and captions, yet little existing research connects video caption evaluation with T2V generation assessment. This paper introduces VidCapBench, a scheme specifically designed for generation, agnostic to any particular format. VidCapBench employs data annotation pipeline, combining expert model labeling human refinement, associate each collected key information spanning aesthetics, content,...
Based on 58×106 J/ψ events collected with the BESII detector at Beijing Electron-Positron Collider, baryon pair processes J/ψ→Σ+Σ¯− and J/ψ→Ξ0Ξ¯0 are observed for first time. The branching fractions measured to be B(J/ψ→Σ+Σ¯−)=(1.50±0.10±0.22)×10−3 B(J/ψ→Ξ0Ξ¯0)=(1.20±0.12±0.21)×10−3, where errors statistical second ones systematic.Received 12 October 2008DOI:https://doi.org/10.1103/PhysRevD.78.092005©2008 American Physical Society
Despite Graph neural networks' significant performance gain over many classic techniques in various graph-related downstream tasks, their successes are restricted shallow models due to over-smoothness and the difficulties of optimizations among other issues. In this paper, alleviate over-smoothing issue, we propose a soft graph normalization method preserve diversities node embeddings prevent indiscrimination possible over-closeness. Combined with residual connections, analyze reason why can...
Despite Graph neural networks' significant performance gain over many classic techniques in various graph-related downstream tasks, their successes are restricted shallow models due to over-smoothness and the difficulties of optimizations among other issues. In this paper, alleviate over-smoothing issue, we propose a soft graph normalization method preserve diversities node embeddings prevent indiscrimination possible over-closeness. Combined with residual connections, analyze reason why can...