Sungsoo Ha
PhD Candidate | Research Assistant
Visual Analytics and Imaging Lab
Computer Science
Stony Brook University, New York
E-MAIL: sunha@cs.stonybrook.edu, hasungsoo@gmail.com
PHONE: +1 631 946 2272
ADDRESS: 329 Old Town, E. Setauket, NY 11733
[RESUME]
Tell Me More

About

Sungsoo Ha

Research Assistant | VAI Lab

I am now a research assistant at Visual Analytics and Imaging Lab, Stony Brook University under the supervision of Prof. Klaus Mueller. My research interests include medical imaging, machine learning, visualization, big data, robotics, and high-performance computing on GPUs. Currently, I am working on three exciting projects:

  • GPU acceleration of statistical iterative CT reconstruction algorithm.
  • Metal artifact reduction in CT data with known metal components.
  • Communication framework of two 6-axis robot arms synchronized by motion capture system.

As a person I am trying to be humble by accepting my limitations, listening others' words, and rediscovering a sense of wonder.

Research

GPU-Accelerated SIR for CT

#SIR #GPU

Low-dose CT with patch DB

#BigData #Patch

Memorable Visualization

#ChartJunk #InfoGraphic

Framework for low-dose CT

#BigData #NonLocalMeans

Data-Driven Color blending

#HuePreserve #Visualization

3D DIRECT TOF PET Recon.

#DIRECT #TOF-PET #GPU

Work Experience

Research Assistant | Visual Analytics and Imaging Lab, Stony Brook University | AUG 2010 - Present

  • Devise an algorithm that improve speed-up more than 100 times in statistical iterative CT reconstruction framework.
  • Develop low-dose CT reconstruction/restoration framework that can reduce radiation exposure amount up to 50%.
  • Develop GPU-accelerated cone-beam CT reconstruction toolkit.
  • Build synchronous communication framework between two 6-axis robot arms using motion capture system.
  • Maintain NVIDIA K20 GPU cluster server.

Software Developer | CEWIT Korea | AUG - NOV 2014

  • Develop 3D transformation of 2d video contents based on quantization method and authoring tool for stereoscopic video.
  • Focus on GPU-acceleration to make real-time smoothing operation of quantized depth map.

Visiting Scholar | Medical Imaging Processing Group, University of Pennsylvania | MAY - AUG 2011, 2012

  • Develop GPU-accelerated 3D DIRECT TOF PET reconstruction framework.
  • Enable to handle spatially varying 3D kernels for forward/backward projection operations.

Teaching Assistant | Department of Computer Science, Stony Brook University | JAN - DEC 2010

  • CSE 102: Introduction to Web Design and Programming
  • CSE 304: Compiler Design