Sehwan Kim Email & Phone Number

Usa

- Develop a VR communication platform, Vizible, for enabling remote users to collaborate in a virtual environment.- Develop algorithms for estimating the pose of WorldViz PPT (Precision Position Tracking) by fusing it with IMU (Inertial Measurement Unit) even when PPT is lost and recovered by using Kalman filter and dead reckoning. - Develop VR Development Kits and Custom software (e.g., real-time panorama display, anatomy simulator, psychology experiments, etc) using WorldViz Vizard (VR Toolkit), C++ and Python. Teach Vizard programming in seminars- Developed a Stereoscopic interactive system of Anatomy which enables users to virtually dissect, navigate and observe the human body structures from any angle in real-time. The system provides several interesting features, such as, sharing the same navigation experience among teachers and students, on-line quiz, video play for each part of the human body, etc.- Developed a VR Development Kit, Wirks (WorldViz Immersive Reality with Kinect and Sony) which enables users to design and build their own Kinect applications for everything from scientific research to entertainment & architectural exploration using Vizard’s straightforward development environment.

Santa Barbara, California Area

- Acquired surround-view panoramas using a single hand-held or head-worn camera based on real-time camera orientation tracking. - Presented methodology for camera orientation relocalization using ‘virtual keyframes’ for online environment map construction when tracking fails.

Santa Barbara, California Area

- Analyzed image feature descriptors (Log-polar Gabor Jet Descriptor and Fourier Descriptor) for a fully automatic image mosaicking of an arbitrary number of images, and implemented an experimental testbed for evaluation.- Developed Advanced Image Pair Registration (AIPR) software that registers unordered sets of images in a completely automatic manner. (Using C#, WPF, .Net Framework)

Santa Barbara, California Area

- Based on the concept of ‘Anywhere Augmentation: Mobile AR in any situation and location’, generated a 3D building model on the fly using an aerial photograph and a video image, and estimated a camera pose using an inertial sensor, a camera and GPS in real-time for virtual text annotations. Investigated sensor fusion based on an Unscented Kalman Filter (UKF) for pose estimation

Korea

- Investigated 3D pose estimation, registration and interaction methods for Mobile Augmented Reality System (MARS) based on the concept of ‘Anywhere Augmentation’

Korea

- Generated MR/AR environments with registration methods for a photo-realistic 3D scene reconstruction.- Investigated multi-modal full-body interaction methods, based on disparity and motion information, in real-time.- Investigated 3D object segmentation/tracking algorithms with asynchronous digital cameras.- Developed 3D vision-based natural interaction methods for AR systems.

Korea

- Investigated image retrieval algorithms, and face detection and facial feature extraction algorithms.- Developed various image/video coding algorithms which include transform, quantization, and entropy coding.- Extensive knowledge on JPEG, MPEG-1, -2, -4, -7, H. 261, H. 263, etc.

Doctor Of Philosophy - Phd, Computer Vision

Activities and Societies: Research Assistant in U-VR Lab., Teaching AssistantDoctoral Dissertation "Indoor Scene 3D Reconstruction.

Summer Session, Computer Science

Master Of Science - Ms, Image Processing

Activities and Societies: Research Assistant in Visual Communication Lab.MS Thesis "A Study on Illumination Adaptive Face Detection.

Bachelor Of Science - Bs, Electronic Engineering