Wontaek Kim Email & Phone Number

Atlanta, Ga, Us

South Korea

• Tested and simulated camera ISP (Image Signal Processing) pipeline code for Galaxy Z Flip/Fold 5 within Qualcomm chipset• Created an ISP simulator using algorithms originally designed for Galaxy device models. Incorporated features such as BPC (Bad Pixel Correction), demosaic, and gamma correction• Visualize the procedure of Galaxy ISP pipeline by converting RAW images into sequentially dumped JPG images• Automated the adjustment of camera parameters for changing user-settings according to surrounding environment which resulted in enhanced user experience via 23% faster load times

Atlanta, Georgia, United States

• Constructed algorithms for automatic counting of patterns (etched fissions) in minerals from microscope 3D images by using Deep Neural Networks (DNN), and integrated in ImageJ plugin to run promptly• Trained two deep neural networks with object detection (YOLOv3) and algorithms (Darknet-53) with real datasets from the lab and connected the algorithms into offline application “AI-Track-tive” by using OpenCV• Automated the enumeration of fissions in target minerals for collecting density data of fissions per dimension

Atlanta, Georgia, United States

• Built an iOS application for aphasia patients. Implemented a drag-and-drop practice session and recording features those record the patients’ ability of correlating and sequencing the given situations so that clinicians can listen to them and interpret the data uploaded on Dropbox• Revised issues and designed a new version of previous Discourse Narrative Story Cards (DNSC) application with new features such as recording indicator and admin pannel