Ziyu Liu Source Confirmed

OverviewAI-generated summary

Ziyu Liu's research focuses on the application of advanced machine learning techniques to diverse analytical challenges. Recent work includes the development of unsupervised methods for classifying plot morphologies using graph attention networks, with applications in historical urban studies. Liu has also investigated ensemble models for differential analysis and explored multiscale information fusion for bacterial detection, drawing inspiration from large language models. Further research extends to reinforcing multimodal generative reward models with agentic tool use and visual reasoning, and developing unified pre-training frameworks for self-attention mechanisms. Liu has also contributed to interactive systems, such as a broadcast gymnastics training system based on pose estimation, and algorithmic approaches for clustering seismic waveforms.

Metrics

• h-index: 1
• Publications: 13
• Citations: 4