Sangmin Oh

Title: Innovations of Sangmin Oh in Causal Analysis

Introduction

Sangmin Oh is an accomplished inventor based in Clifton Park, NY (US). He has made significant contributions to the field of causal analysis through his innovative patent. His work focuses on improving the retrieval and classification of causal sets of events from unstructured signals.

Latest Patents

Sangmin Oh holds a patent titled "Systems and methods for retrieving causal sets of events from unstructured signals." This patent describes a method that enhances performance in retrieving and classifying causal sets of events from unstructured signals. The method involves applying a temporal-causal analysis to the unstructured signal, which includes representing the occurrence times of visual events as a set of point processes. An exemplary embodiment interprets visual codewords produced by a space-time-dictionary representation of the unstructured video sequence as point processes. A nonparametric estimate of the cross-spectrum between pairs of point processes is obtained, and a spectral version of the pairwise test for Granger causality is applied to identify interaction patterns between visual codewords. This innovative approach allows for the grouping of these interactions into semantically meaningful independent causal sets. Furthermore, the method leverages segmentation achieved during temporal causal analysis to enhance the categorization of causal sets.

Career Highlights

Sangmin Oh is associated with the Georgia Tech Research Corporation, where he continues to advance his research and innovations. His work has garnered attention for its potential applications in various fields, including video analysis and machine learning.

Collaborations

Sangmin has collaborated with notable colleagues such as James Rehg and Karthir Prabhakar, contributing to a dynamic research environment that fosters innovation and discovery.

Conclusion

Sangmin Oh's contributions to the field of causal analysis through his patent demonstrate his commitment to advancing technology and understanding of unstructured signals. His innovative methods have the potential to significantly impact various applications in the future.