Research

Kihwan Kim

Kihwan Kim, Ph.D.
Senior Research Scientist
Kihwan Kim's picture
Bio:

Kihwan Kim received Ph.D degree in Computer Science from Georgia Institute of Technology in 2011, and BS from Yonsei University in 2001. Prior to join Georgia Tech, he spent five years as an R&D engineer at Samsung and also worked for Disney Research Pittsburgh as a visiting research associate/research intern for 8 months during his graduate study.

His research interests span the areas of computer vision, graphics, intelligent systems and multimedia. A common thread in his research is in understanding dynamic scenes from videos, and visualizing the motion and structure extracted from the scene.  Currently, he is leading NVIDIA's SLAM projects, and also working on gesture recogntion system based on different modalities (sensors).

For a complete list of papers, including those published before joining NVIDIA research, see here

Research Interests:

Motion and Video Analysis, Object detection/tracking, Stereo and Structure from motion, MVS/SLAM, Numerical method and optimization, Real-time rendering, Scattered data approximation, Augmented reality, etc.

Publications:
Reflectance Estimation On-The-Fly
Towards Selecting Robust Hand Gestures for Automotive Interfaces
Accelerated Generative Models
Online Detection and Classification of Dynamic Hand Gestures with Recurrent 3D Convolutional Neural Networks
MLMD: Maximum Likelihood Mixture Decoupling for Fast and Accurate Point Cloud Registration
Filtering Environment Illumination for Interactive Physically-Based Rendering in Mixed Reality
Hand Gesture Recognition with 3D Convolutional Neural Networks
Short-Range FMCW Monopulse Radar for Hand-Gesture Sensing
Multi-sensor System for Driver’s Hand-Gesture Recognition
DT-SLAM: Deferred Triangulation for Robust SLAM
WYSIWYG Computational Photography via Viewfinder Editing
Detecting Regions of Interest in Dynamic Scenes with Camera Motions
Gaussian Process Regression Flow for Analysis of Motion Trajectories
Augmenting Aerial Earth Maps with Dynamic Information
Motion Field to Predict Play Evolution in Dynamic Sport Scenes