Information Technology Engineer

Kumar Srijan

Centre for Visual Information Technology

Department of Computer Science and Engineering Ph:+91-996*-***-***

IIIT Hyderabad http://researchweb.iiit.ac.in/ kumarsrijan

OBH 133, IIIT Hyderabad, Hyderabad 500032 email: *****.******@********.****.**.**

Education

IIIT Hyderabad, India 2006 - Present

B.Tech + MS (By Research) in Computer Science and Engineering

CGPA: 7.93/10

Advisor: Prof. C. V. Jawahar

Publications

Kumar Srijan, Syed Ahsan Ishtiaque, Sudipta N. Sinha, C.V. Jawahar. Image-based

walkthroughs from partial and incremental scene reconstructions, Oral Presentation,

In Proceedings of the British Machine Vision Conference(BMVC 2010), Aberystwyth, UK,

August-September 2010

Work Experience

Research Assistant at Center for Visual

Information Technology, IIIT Hyderabad August 2010 - Present

Ongoing work on E cient Match Graphs for Large Image Collections.

Software and Systems Engineer for Virtual Labs

at IIIT-Hyderabad June 2010 - Present

Worked on the end-to-end development and systems management of two Virtual Labs aimed at

helping students learn Computer Programming and Problem Solving.

Systems Admininstrator at Center for Visual

Information Technology, IIIT Hyderabad February 2011 - Present

Work involved administration of computational servers and a cluster of multicore computers.

Teaching Assistant for Software Systems and

Design(SSAD) and Project under Prof. Kirti Garg August 2009 - Decemeber 2009

Work included teaching, mentoring and evalutation of projects, grading assignments and answer

sheets.

Summer Intern at IIIT Hyderabad Summer 2009

Worked on developing a virtual tour of Golkonda fort at Hyderabad.

Member of Video, Design and Literary Team of

Felicity 09, the annual fest of IIIT Hyderabad August 2008 - February 2009

Contributed to the design and creation of the video, logo, website and brochure for Felicity 09.

Organized two literary events in Felicity 09.

Summer Intern at IIIT Hyderabad Summer 2008

Developed a system to capture the 3D motion of simple objects in realtime using stereo scopic vision.

Academic Achievements

Recipient of Undergraduate Research Award for the year 2009-2010 at IIIT Hyderabad.

Included in the Dean s List for academic excellence for the year 2008-2009 at IIIT Hyderabad.

Secured All India Rank 1917 among more than 800,000 candidates in All India Engineering

Entrance Examination, 2006.

Secured All India Rank 3354 among more than 500,000 candidates in IIT Joint Entrance Exam,

2006.

Major Projects

E cient Match Graphs for Large Image Collections (Ongoing Masters Thesis)

In this project, we explore methods for extracting geometric and appearance based similarities

between images to build e cient match graphs which can be used for purposes like Walkthroughs,

Image retrival, Model reconstruction and Image-based Navigation.

Large scale image window clustering for fast object detection on multi core architectures

Developed an iterative sample-build-populate algorithm for clustering a large number of query

windows for running on mulicore architecture for doing object detection in large datasets.

Image-based walkthroughs from incremental and partial scene reconstructions

We built a incremental system for creating walkthroughs by employing a series of overlapping partial

scene reconstructions which completly avoid the drift problems and scalability bottlenecks faced by

traditional methods employing incremental SFM.

Analysis of the Performance of Various Loss Functions in SVM

Implemented the ramp loss function for SVM, using the CCCP procedure to do the non convex

optimization, to demonstrate thier better suitability to datasets involving noise points and outliers.

Capturing, Processing and Experiencing Indian Monuments (Undergraduate Thesis)

Developed a pipeline for visualizing a scene using multiple 2 frame reconstructions from the given

images.

Multiple Kernel Learning for Category Recognition

Ran the MKL-Manik learn SVM code for object recogniton on a new self-created dataset comprising

of new categories of images and did implemetation for the introduction of a new feature to the

MKL-based object recognition pipeline.

Minor Projects

Image Fusion for Context Enhancement

Implemented a paper based on gradient domain techniques to enhance the illumination

characteristics of a night time image/video using the corresponding context from a day time video.

Rotation invariant Pattern Recognition using Zernike Moments

Used Zernike and Geometrical moments as features to recognize characters and digits using a Neural

Network based classi er.

Information Extraction in cricket videos

Extacted information like crowd detection, pitch detection, score detection, ball detection etc. using

computer vision techniques on cricket videos.

Voting solution for Mobile phones

Designed a robust pipeline to address the security issues in Mobile Phone for casting votes by using a

framework to split data into chunks during encryption and storing them on di erent servers.

Tracking motion in 3D

Using a pair of webcams, Implemented a simple stereo scopic vision based 3D motion tracker and

visualizer for simple objects in OpenCV and OpenGL.

Creation of Felicity 09 video

Work involved using of advanced video and image editing softwares like Adobe AfterE ects and

Adobe Photoshop.

Skill Set

Operating Systems: Windows, Linux

Programming Languages: C, C++

Scripting Languages: Bash, Python, PHP, Matlab

Libraries and APIs: OpenGL, OpenCV, MySQL

Web-Based Technologies: HTML, CSS, Javascript, Web2py

Conferences Attended

British Machine vision Conference(BMVC) 2010, Aberystwyth, UK

Microsoft Research India Vision Shindig, 2010

Contact this candidate