
VA System and Method
Conducted machine learning engineering to train, update and deploy object detection models in a web service backend to test and verify the effectiveness of VATLD: a visual analytics system to assess, understand and improve traffic light detection. The work contributed to research paper [J2], which won IEEE VIS 2020 best paper award. Patent granted or under application: US Patent No. 11,301,724 [P2], US Patent App. 17/186,392 [P3]
Updated the framework and web service backend to support the training, updating and deployment of semseg models to test and verify the effectiveness of visual analytics approach for diagnosing and improving semantic segmentation of movable objects. The work contributed to research paper [J1], which was accepted by IEEE VIS 2021 and published in IEEE Transactions on Visualization and Computer Graphics, vol. 28, no. 1, pp. 1040-1050, Jan. 2022. Patent granted: US Patent No. 11,803,616 [P1].