Sept 2010 - Sept 2011 (Seattle Children's
Hospital Research Institute)
|
Postdoctoral Fellow in the
Center of Tissue and Cell Sciences and Center of Child Health, Behavior
and Development.
1. POP Project
2. ILP Project:
- pre and post-op shape change
- shape descriptors for unicoronals
4. Autism 3D
5. CIRAS
- Writing scripts to automate process
6. 3DMD Craniosynostosis
|
|
| Sept
2006 - August 2010
(University of Washington) |
Summer 2010:
Break before starting work :) Holiday to Jakarta and Singapore.
Kiran started daycare
Spring 2010:
TA for CSE576
Thesis defense and graduate!
Winter 2010:
Review for Pattern Recognition journal.
Working on the FaceBase project FaceBase link Project
link
Have a baby! (Update: Kiran Atmosukarto Gomulya was born on Tuesday, 19
January 2010 at 16:22 weighing 6lbs 10.1oz and 19.5in)
Finish writing thesis!
Present oral paper at ACM Multimedia Information Retrieval March 29-31,
2010 in Philadelphia.
Fall 2009:
Learning approach for various craniofacial disorder classification,
trying to use genetic programming to develop quantification
T1 plagiocephaly paper on correlation between severity scores and
cognitive behavior.
Started writing thesis!
Submitting paper to ICPR: genetic programming and Melissa's 2D
brachycephaly measurements- Accepted April 2010.
T2 plagiocephaly paper on shape changes between T1 and T2 data.
Summer 2009:
Revising journal paper for Pattern Recognition (Submitted August 2009)
- accepted October 2009.
Working on learning approach for craniofacial disorder classification
and new
brachycephaly measures.
Mentoring summer student Melissa Gross to work on brachycephaly
classification, quantification and localization.
Working on ICIAP slides - Sept 2009.
Submitted paper to Grail retreat. pattern recognition paper
Submitted paper to MIR 2010: LFD Salient views - accepted November 2009
Spring 2009:
Writing a short journal paper for Pattern Recognition Letters
(Submitted April
2009) - rejected Sept 2009.
Learning features to classify various craniofacial conditions (start
May 2009)
Winter 2009:
Reviewer for IJCV special issue on 3D Object Retrieval.
Presenting at Google Workshop for Women Engineer in Mountain View -
January 2009.
Presenting poster for SPIE Medical Imaging in Florida - February 2009.
Reviewer for Workshop 3D Object Retrieval, Eurographics 2009.
Assisted in writing a NIH grant on midface hypoplasia - Update Sept
2009, grant was accepted, project
webpage.
Submitted a paper on "Shape-based classification of 3D head data"
to ICIAP 2009 - accepted.
Submitted a journal paper to Patter Recognition (February 2009) -
revise and resubmit June 2009.
Submitted a medical journal paper to American Cleft Palate Craniofacial
journal (March 2009) - accepted May 2009.
Autumn 2008:
Attending the computer vision reading group.
Presented poster at the Affiliates Meeting.
Presented our poster at ACM Multimedia Information Retrieval in
Vancouver - October 2008.
Presented my thesis proposal - November 2008
Presented our paper at S+SSPR 2008 in Florida - December 2008.
Summer 2008:
Worked on my general exam - thesis proposal paper.
Spring 2008:
I presented our poster at Eurographics 2008 Workshop on 3D Object
Retrieval in Crete, Greece.
I presented our multimedia paper at the SIIM Meeting in Seattle.
I volunteered at CVPR 2008 in Alaska
Winter 2008:
I am TA-ing for the professional masters computer vision class
(CSEP576).
Reviewer for CIVR 2008
Fall 2007:
I am co-organizing a computer vision reading group (cse590v)
in the department. The aim is basically to read some of the recent
vision papers that we think are interesting and have significant impact
in the vision community.
I am also taking one more post-quals class from the Biomedical Health
Informatics department MEBI534.
In October 2007, I attended the Grace Hopper Conference in Orlando,
Florida.
I made a poster for the Affiliates Meeting
2007 in the department (ppt).
Summer 2007:
Research Assistant for the Multimedia Grant.
Spring 2007:
I am TA-ing the graduate class for computer vision (CSE576).
The class is thought by Prof Linda Shapiro.
Attended CVPR in Minneapolis.
Fall 2006 and Winter 2007:
Develop the initial web-accessible prototype for the Multimedia
Information Retrieval grant.
Attended Grace Hopper Conference in San Diego.
|
I am part of the 'Multimedia
Information Retrieval for Biological Research' grant. The aim is to
develop a probabilistic unifying query framework for
retrieval of
biological data of different modalities such as 1D time series signals,
2D images, 3D image volumes, text data and 2D and 3D images over time.
This would involve developing features and similarity measures for each
of the biological data types and finding the best way to combine these
different features and similarity measures. I developed the first prototype system
for the three motivating
applications: brain data retrieval for study of language sites in the
brain, skull data retrieval for studying the effects of
craniosynostosis, and mouse eye image retrieval for the study of
cataract development. Here
is the poster we presented at the ACM Research Night in October 2006.
My research focus is in computer
vision and right now I am focusing on learning what are the best
descriptor and
similarity measures for 3D data according to the particular task given
such as
retrieval and classification. I am also working with clinicians at the
Children's Hospital to find ways to quantify the severity of skull
flatness in children with plagiocephaly. |
|
Conferences to look out
for: CVPR, ICCV, ICIP, CVIR, MICCAI, ECCV, MM, SIGIR, ICME, AMIA, ICPR,
WACV, CAIP
|
Journals to look out for:
PAMI,
IJCV, CVIU, ACM Transaction on Graphics, International Journal of Shape
Modeling
|
|
|
Summer
2006 (Siemens Medical
Solutions)
|
I did a
summer internship at Siemens Medical Solution, Malvern, PA with the
Computer Aided Diagonosis (CAD) group from June to end of August 2006.
My
project was on automatic lymph node detection on CT scans which
involved finding the best features to represent a lymph node in CT
scans, for eg. curvature, gradient etc (Image Processing).
|
|
| April
2005 - June 2006
(University of Washington) |
Starting from
Spring quarter I will be working on a new project in collaboration with
the HITLab at University
of Washington. The project is related to medical vision using an
endoscope.
Project Description: a system is being developed to assist
bronchoscopist in navigating an ultrathin endoscope through the
bronchial using an electromagnetic tracking device and virtual
bronchoscopy. A 3D reconstructed model of the bronchial tree is
obtained through segmentation of a high resolution CT scan. The
electromagnetic sensor is positioned at the tip of the endoscope and
provides information on the top position and orientation in the
endoscope a in order to produce real-time virtual fly-through
perspective of the lung at that position.
Quals Project:
Presentation
slides and the videos
Quals
paper - Interactive 3D User Interface for Bronchosopy
Course Projects:
CSE576 Computer Vision Project 1 - Feature Detection and Matching (report)
CSE576 Computer Vision Project 2 - Panoramic Mosaic Stitching (report)
CSE576 Computer Vision Project 3 - Fully Automated Panoramic Mosaic
Stitching (report)
CSE576 Computer Vision Project 4 - 3D Reconstruction of Antique Stereo
Pairs (report)
CSE577 Advanced Computer Vision Project - Camera Tracking for
Bronchoscopy (report)
CSE561 Computer Communications and Networking - Build wireless adhoc
network using sound waves as the means of communication
CSE548 Computer Architecture - Optimizing Dijkstra's Algorithm on
Dataflow Architecture. Used plenty of assembly language on WaveScalar
for this!
CSE503 Software Engineering - term paper on Automated Theorem Proving
and tool evaluation on ESC/Java2
Activities:
Reviewer for CIVR 2006 (Conference on Image and Video Retrieval)
|
|
| Sept
2004 - March 2005
(University of Washington) |
|
Presently I am working on the Video Analysis and Content
Exploitation Project (VACE) under ARDA. I am taking over the system
built by Yi Li. I will be maintaining it and adding on extra features
as the research progresses. The objective of the project is to
recognize all objects in each frame of the video , track the objects,
and to detect interesting events such as cars overtaking other cars etc
by continuously tracking the objects in each frame.
Course Projects:
CSE573 Artificial Intelligence Mini Project I - Build two
satisfiability solvers (DPLL & WalkSAT) (report)
CSE573 Artificial Intelligence Mini Project II - SPAM Filter (report)
Attended CRA-W in San Fransisco.
|
| |
| 2001
- 2004 (National
University of Singapore) |
|
I previously worked on a 3D retrieval
project under A/P Leow Wee Kheng and I helped out on the DIVA
(Digital Image and Video Album) project.
Here is a demo video of how the 3D
retrieval system works. The user can basically specify positive and
negative examples to constraint the search results.
In addition, I was starting out in a Protein Docking project where the
objective of the project is to determine the matching parts of proteins
based on their 3D structures. This work is related to the 3D Retrieval
project as we could use the prior results to determine useful 3D
features for the protein.
|
| |
| Masters
(2000 - 2001)
(National University of Singapore) |
| Project title:
Frontier
Advancing Delaunay Triangulation of unorganized 3D points. |
| Supervisor: A/P
Leow Wee Kheng |
| Abstract: |
|
Mesh
construction is an important stage in the acquisition of 3D object
models. It constructs triangle meshes based on the 3D points recovered
from image sequences, laser scanning or other recovery techniques to
represent the surface shapes of the objects. This thesis describes a
novel frontier advancing polygonization algorithm for constructing
triangle mesh that represents the surface of the objects from
unorganized 3D points. It produces 2D triangles that approximate the
surfaces of the 3D model. In contrast to existing methods, this
polygonization algorithm begins the mesh construction process from
reliable 3D points lying on smooth surfaces, and advances the mesh
frontier towards ambiguous 3D points lying near surface
discontinuities. It is shown that this frontier advancing algorithm
performs 2D Delaunay triangulation of 3D points lying on a plane in 3D
space. For 3D points lying on a smooth curved surface, the constructed
mesh is a Delaunay triangulation of a piecewise planar approximation of
the curved surface.
If you are
really interested, you can download my thesis.
|
| |
| Honours
(1999 - 2000)
(National University of Singapore) |
| Project title:
Polygonization of
Noisy 3D Surface Points |
| Supervisor: A/P
Leow Wee Kheng |
Abstract:
|
|
The
objective of this project is to reconstruct the 3D model of an object
from a set of 3D points. Most of these 3D points correspond to feature
points on the object's surface. However, due to finite camera
resolution and the presence of image noise, some of these 3D points are
noise points which do not correspond to any feature points on the
object's surface. The project objective is achieved by firstly removing
the noise points from the 3D models. Two algorithms are implemented to
accomplish this step. The first method implements a clustering
algorithm to identify the noise points. This method is able to
recognize noise points that are far from the object's surface. The
second method applies the epipolar constraint principle to identify
noise points that are close to the object's surface. Hence, the
combination of these two methods result in a less noisy model. After
removing the noise points, the surface of the model is built using a
new polygonization algorithm. Unlike existing polygonization algorithms
such as the Alpha Shape algorithm, the new algorithm can detect
possible surface discontinuities of surfaces and reconstruct these
surface discontinuities.
A copy of my
honours thesis is available in SOC digital
library.
You can also
download my honours
presentation slides.
|
| |