Prachi Agrawal
From Msplacements
-
E-mail: prachi@research.iiit.ac.in, prachi.agrawal@gmail.com
- Ongoing MS by Research (Computer Vision), IIIT-Hyderabad, (expected December, 2009);
- B. Tech. (ECE), IIIT- Hyderabad, June 2008, CGPA 8.85;
- Senior Secondary (CBSE), K. C. H. S. S. School, AISSCE - 2003, 87%;
- Secondary (CBSE), St. John’s School, AISSCE - 2001, 93.8%
-
Research Assistant, CVIT, May 2008 - till date.
Project: A Real-Time system for Human De-identification in Surveillance Videos. -
Teaching Assistant, Spring 2008.
Computer Graphics. -
Teaching Assistant, Monsoon 2007.
Electrical Science. -
Student Intern, CVIT, May 2006 - Jul 2006.
Project: OSG implementation of huge models for faster and efficient rendering on a 4x4 Display Wall. - Have been awarded a scholarship by IARCS to attend ACCV '09 to be held in Xi’an, China.
- Figured in the Dean's List for Academic Excellence in Spring 2005, Spring 2006, Spring and Monsoon 2007, and Spring 2008.
- Figured in the Merit List for Academic Excellence in Monsoon 2004 and Monsoon 2006.
- Best circuit award for Mini-TT in the Electronic Workshop open house in the 3rd semester (2005-06).
- Successfully completed an Elementary level course in German, organized by Goethe-Zentrum, in June 2008.
- Was one among a small batch of students selected to attend Dr. A.P.J. Abdul Kalam's lecture series in IIIT-H.
- Was the topper of my city, Firozabad, in CBSE - AISSCE-2001 (X), Secondary Examination.
- Going to attend and present my work in the Ninth Asian Conference on Computer Vision (ACCV 2009), to be held in Xi’an, China (Sep. 2009).
- Attended Indian Conference on Vision Graphics and Image Processing (ICVGIP 2008), held in Bhubaneswar, India (Dec. 2008)
- Attended TechVista, Microsoft Research India's annual research symposium, held in Chennai, India (Oct. 2008).
- Attended the first Indo-Israel Workshop on Computer Vision (WCV 2008), held in Hyderabad, India (Feb. 2008).
- We outlined the scenarios in which de-identification is required and the issues brought out by those.
- We presented a preliminary approach to de-identify individuals from videos. A bounding box around each individual present in a video is tracked through the video. An outline of the individuals is approximated by carrying out segmentation on a 3-D Graph of space-time voxels. We explored two de-identification transformations: exponential space-time blur and line integral convolution.
- We collected videos in a plausible setting and conducted a detailed user-study to validate the effectiveness of the de-identification schemes.
-
Video Object Cut and Paste
Faculty Advisor: Dr. P. J. Narayanan
Honours Project, CVIT, IIIT-H.
The aim of the project was to segment out objects from a video in near real time using Graph Cut Segmentation in 3D, where graphs were built on voxels (superpixels) rather than pixels. The segmentation for the keyframes was provided by the user and the foreground and background information was propogated across the 3D graph using Graph Cut.
Technologies used: C++, C, Matlab -
OSG Implementation of Huge Models for Faster and Efficient Rendering on a 4x4 Display Wall
Faculty Advisor: Dr. P. J. Narayanan
Summer Project, CVIT, IIIT-H.
The aim of this project was to create Open Scene Graph models for heavy models of the Fatehpur Sikri (14,000 objects) and a Power Plant (1,000+ objects). The purpose was to exploit the hierarchy within the models and use the inherent property of hierarchical culling of OSG for efficient rendering on a 4x4 Display Wall.
Technologies used: OSG -
Adaptive GMMs based Background Subtraction on GPU
Faculty Advisor: Dr. C. V. Jawahar
Semester Project for the course Computer Vision
The aim of this project was to develop a fast video processing library on GPU using CUDA API. We implemented the background subtraction algorithm which was based on the adaptive Gaussian Mixture Models for background learning. We showed an improvement between 500% to 1900% over the C++ implementation.
Technologies used: C++, CUDA. -
Color Based Image Retrieval
Faculty Advisor: Dr. Anoop Namboodiri
Semester Project for the course Digital Image Processing
The aim of this project was to achieve image retrieval based on color content through implementation of a KLT based histogram space compression for faster calculation.
Technologies used: C, Matlab -
Mini-TT
Faculty Advisor: Dr. Jayanthi Sivaswamy
Semester Project for the course Electronic Workshop.
The aim of this project was to design and build a ping-pong game for one/two players on the circuit board. A series of consequently glowing LEDs were used to show the motion of the ball and switches were used to implement virtual bats for the players.
Circuit elements used: NC 555 chip, variable resistor, Three way switches, LEDs. - MC6802 implementation on VC++.
- 2D real-time implementation of the popular game TETRIS.
- 3D simulation of a train.
- DES Encryption/Decryption algorithm.
- Count-down Timer with an Buzzer.
- Temperature Sensor.
- Rudi Cilibrasi and Paul Vitanyi’s paper on “Automatic Meaning Discovery using Google”
- Study of Modeling Speech Signals
- Study of various Encryption/Decryption algorithms
- Programming Languages :- C, C++, CUDA
- Scripting Languages : - Python, Shell, Matlab.
- Operating Systems : - GNU/Linux, Windows 95/98/2000/XP.
- Web Languages : - HTML, XML.
- Programming Environments : - GNU/GCC, Visual Studio
- Libraries : - OpenCV, Newmat, ffmpeg, OpenGL, OpenSceneGraph, FANN, Qt.
- Tools : - MS Office, Photoshop/Gimp.
- Publishing : - LaTeXe, LEd
- Was awarded a commendation certificate for "outstanding contributions towards organizing events in cultural and sports arena" during my 4 year stay in my college, IIIT Hyderabad.
- Was the Sponsorship Head of Felicity 07, the annual cultural and technical fest of IIIT-H. Was awarded an appreciation certificate for "extra-ordinary efforts".
- Was the Cultural House Secretary in 2007-2008, the Captain of Agni House in 2006-2007, and a Member of the Cultural Council for two years.
- Was a Member of Students Parliament of IIIT-H for three years.
- Was a Member of the Student Mentor Body which was recruited to monitor the interaction between the seniors and the freshmen during the academic year 2007-2008.
- Was the co-founder and the co-ordinator of the Dance Club in IIIT-H in the academic year 2006-2007.
- My team won Inter College Western Group Dance Competitions in Felicity 05, Felicity 06 and Felicity 07.
- Won 2nd prize in Inter College Two-to-Tango Competitions for duet performances in Felicity 08 and Felicity 09.
- Won various Inter House Dancing and Singing Competitions in both, solo and group, categories.
- Was a member of the House and College throwball team.
- Was the Deputy Headgirl and the House Captain in school. Won many solo singing, declamation, skit, essay writing competitions in school.
- Secured 10th position out of 1200 participants, in the stock market simulation game, Equity Pulse, organized at Threads, in IIIT in 2006.
Contact No: +91 9866217459
Career Objective
Challenges inspire me and I am driven by an urge to achieve perfection in everything I take up. My objective is to apply my academic experience to the latest research going on in the field of Computer Vision and Image Processing.
Education
Work Experience
Academic Achievements
Publication
Prachi Agrawal and P. J. Narayanan, “Person De-identification in Videos”, The Ninth Asian Conference on Computer Vision (ACCV 2009), Xi’an, China, Sep. 2009.
Conferences and Workshops
Primary Research
Person De-identification in Videos
(Advisor: Dr. P. J. Narayanan)
-
Advances in cameras and web technology have made it easy to capture and share large amounts of video data over to a large number of people through services like Google Street View, EveryScape, etc. A large number of cameras oversee public and semi-public spaces today. These raise concerns on the unintentional and unwarranted invasion of the privacy of individuals caught in the videos. To address these concerns, automated methods to de-identify individuals in these videos are necessary. De-identification does not aim at destroying all information involving the individuals. Its goals are to obscure the identity of the actor without obscuring the action. In this project,
Major Projects
Other Projects
Studies
Skillset
Advanced Courses Taken
Pattern Recognition, Computer Vision, Computer Graphics, Digital Image Processing, Cryptography and Network Security, Data Compression, Spatial Informatics, Linear Programming.
Academic Interests
Biometrics, Pattern Recognition, Computer Vision, Image Processing, Data Structures, Cryptography and Data Compression.
Extra Curricular
