Deep-Learning Based Approach for Scene Analysis and Trajectory Tracking

Project Details
STATUS

In-Progress

START DATE

10/01/18

END DATE

07/31/19

SPONSORS

SMART AG LLC

Researchers
Principal Investigator
Anuj Sharma

Co-Director, REACTOR

About the research

The growing availability of big data, along with computer hardware improvements, especially GPUs, attracts researchers and engineers to the emerging area of deep learning. One of the main applications of deep learning is its great capability to solve object classification and detection problems. It is demonstrated that deep nets can even outperform humans in these types of problems. Similar to most engineering fields, construction engineering has been also dealing with object detection issues. In this project, the research team will be comprehensively investigating construction workforce detection from development to production. Additionally, this research can be easily generalized to other domains and industries like agriculture because while the data might be different to some extent, the underling concepts and implementations are the same.

TOP