Kyunghwan Jeong

• Formulated an algorithm to assess the potential error rate in a 3D Surround View System image. Introducing this algorithm, which leverages virtual camera pose estimation, significantly optimized the testing process. Testers were able to efficiently evaluate the system's performance by focusing on images with potential errors not every other images, saving both time and resources.
• Identified new problem areas and researched technical details to build innovative solution.
• Worked side by side with colleagues and teams to problem solve.

• Improved a normalize flow-based network, a type of generative model. Implemented an auxiliary loss function based on scale-invariant Signal-to-Distortion Ratio (SDR), addressing the limitation of the existing cost function that solely relied on probability density functions and didn't consider real data. This modification enhances the model's ability to incorporate real-world data.

• Identified, analyzed, and interpreted patterns in global datasets of CT machine malfunctions by establishing correlations and creating a dashboard with charts. Received commendation from the project manager for the insightful effectiveness of the visualization.
• Collaborated with stakeholders in bi-weekly meeting to identify business needs and data sources.

• Created a MATLAB-based Graphical User Interface (GUI) for a multichannel signal enhancement algorithm.
• Utilized PyTorch to assess FeatureNet's performance in choosing optimal microphones for Wide Area Sensor Network (WASN).
• Implemented speech enhancement algorithms into the "Pepper" robot using Python.
• Worked both independently and collaboratively in fast-paced laboratory environment.
• Recorded experimental findings in laboratory notebook to analyze and interpret experimental findings.

GPA: 2.1/1.0

GPA: 3.95/4.50

• Recipient of the Academic Excellence Scholarship on three separate occasions.