Papers
2
Total Citations
5
H-Index
2
About
Qirong Ho is a researcher whose work spans computer vision, large-scale data analysis, and industrial applications of machine learning. His early contributions include the development of "Region Graph Spectra as Geometric Global Image Features" (2009), a method that uses spectral graph theory to capture global geometric properties of images, earning 3 citations and laying groundwork in visual feature extraction. Ho’s most impactful work, "ConTrack: A Scalable Method for Tracking Multiple Concepts in Large Scale Multidimensional Data" (2016), addresses the challenge of analyzing vast, unlabeled temporal datasets from domains like finance, telecommunications, and sensor monitoring. ConTrack enables efficient tracking of multiple evolving concepts—such as user activity patterns or financial transaction anomalies—without requiring labeled data, achieving 2 citations and demonstrating practical utility in real-world industrial settings. By bridging theoretical innovation with scalable, unsupervised solutions for high-dimensional temporal data, Ho has made notable contributions to both academic research and applied data science, offering tools that help organizations extract actionable insights from complex, continuously generated streams of information.
Research Focus
Key Achievements
Top Papers
- 1Region Graph Spectra as Geometric Global Image Features3 citations · 2009
- 2