Engineer System

SOFTWARE DEVELOPER

Toyota Research Engineer

Innovative and versatile professional with over 4 years experience in

developing software and algorithms for real-world problems. Demonstrated

excellent interpersonal, leadership, problem solving, and technical

skills. Proficiencies include:

. Object oriented cross-platform software development (C++, Qt) .

Multithreaded GUIs

. Software analysis and design . CCD and time-of-flight imaging.

. Generic Programming. . Machine learning algorithms.

TECHNICAL SKILLS

Programming languages:

o C++, Matlab, Python, Java, SQL, Visual Basic.

Libraries, IDEs & tools:

o Qt, IPC, PyQt, STL, OpenGL, OpenCV, Swing, Swig, Blender, VCS,

SVN.

o MS Visual Studio, KDevelop, Eclipse, gcc, Netbeans

PROFESSIONAL EXPERIENCE

TOYOTA TECHNICAL CENTER, (Cambridge & Ann Arbor) [10/07 -

Present]

Technical Research Department: Research Engineer:

o

Tasks:

o Developed interface modules based on IPC library which are used by

all processes in the current system for communicating and passing

data. This design makes the system more modular and robust to

crashes.

o Developed a multithreaded cross-platform GUI to capture and

visualize real-time 2D and range data. The GUI displays image and

point-cloud data; renders 3D models; and provides sensor specific

controls. The GUI is the main interface for the whole project and

is currently used for capturing and processing data from 8

different sensors over the network at 60Hz.

o Implemented modules for estimating and rendering 3D human body

pose using articulated models on range (point cloud) data.

o Implemented modules for predicting temporal events (such as

driving, texting and drinking) by matching 3D exemplars using

background subtraction and nearest neighbor algorithm.

o Implemented a supervised learning algorithm called temporal random

forests for analyzing temporal data and evaluated against other

algorithms such as HMM, CRF and HCRF. Developed a GUI for training

and evaluating these algorithms. A variant of the developed

algorithm is currently being investigated for a patent

possibility.

o Integrated various sensors and algorithms from other developers

into a single system by developing generic interfaces, wrappers

and python-bindings.

o Deployed a cross-platform SVN based software development system.

o Imported a windows-based system to Linux and implemented recursive

makefile architecture to compile and build software efficiently.

Selected Achievements:

o Developed and deployed a complete system with sensors and software

on a prototype vehicle. The system provides a user-friendly GUI to

collect data, test algorithms, predict temporal events, assess

risks and provide alerts.

o Analyzed previous system using activity & sequence diagrams.

Proposed and implemented a new design making the system modular,

robust and computationally optimized.

o Documented code, presented demos, proposed sub-projects, published

papers and applied for patents. (Refer to 'Other achievements'

section).

o Demonstrated excellent teamwork, communication and presentation

skills.

Toyota Technical Center, Cambridge, MA. [06/06 -

01/07]

Intern

o Developed tools and Qt based GUI for data collection, training and

analysis.

o Implemented modules to perform face detection & recognition tasks.

Performed data clustering and PCA analysis on feature vectors.

o Implemented modules to train a generic Haar feature classifier on

any object.

Environment & libraries: C++,

OpenCV, Qt.

George Mason University, Fairfax, VA. [01/07 -

06/07]

Graduate Assistant

o Tutored and assisted students in getting hands on experience with

object-oriented programming, data structures and graphical user

interfaces.

o Course topics include trees, heaps, complexity analysis and

recursive algorithms.

Environment & libraries: Java.

George Mason University, Fairfax, VA. Vision and Robotics Lab [10/05 -

05/06]

Research Assistant

o Worked on the problem of learning 'part based' representations of

objects and object categories. Implemented a new feature selection

model called 'strangeness measure' to learn discriminative

features and classify objects.

o Analyzed performance of the model when combined with techniques

such as PCA and K-means clustering. Environment &

libraries: MATLAB, C++, Qt.

ACADEMIC Projects

Intrinsic Colorization from web-based images:

o Implemented an algorithm to colorize grayscale images based on

reference images obtained from the internet as described in the

paper 'Intrinsic Colorization' by Liu, Wan Et al.

o The process involves image registration using SIFT, recovering

intrinsic image components (i.e. illumination and reflectance) and

image colorization.

Environment & libraries: C++,

OpenCV, SIFT.

Rendering volumetric data and silhouettes:

o As a part of the project I've implemented modules to read and

render poly (volumetric) data with feature enhancements. Also

implemented algorithm to render polygonal silhouettes of

volumetric data as described in the paper 'Image Precision

Silhouette Edges'.

Environment & libraries: C++,

OpenGL.

Robot Path Planning Project:

o Implemented A* algorithm to solve the problem of robot navigation.

Developed Swing based GUI to dynamically change robot and goal

positions and also display all possible and optimal paths.

Environment & libraries: Java, Netbeans.

Object Recognition from Cluttered Scenes:

o Developed a system to recognize specific objects from cluttered

scenes using SIFT features.

Environment & libraries: C++,

OpenCV, SIFT.

EDUCATION

. M.S in Computer Science, George Mason University, VA, [08/2005 -

08/2007]. G.P.A - 3.64

. B.E Computer Science, M.V.S.R Engg. College, Hyd. India, [06/2001-

06/2005]

OTHER ACHEIVEMENTS

Publications:

o Recognizing events with temporal random forests. [Co-Author]

ICMI 2009

o Recognizing Gestures for Virtual and Real World Interaction. [Co-

Author] ICVS 2009

Patent Applications:

o 201******** - Electronic control system, electronic control unit

and associated methodology of adapting a vehicle system based on

visually detected vehicle occupant information.

o 201******** - System, apparatus and associated methodology for

interactively monitoring and reducing driver drowsiness.

o 200******** - System and method for reducing boredom while

driving.

Chenna Krishna Reddy Varri

407 S. Fifth Ave., Room 2, Ann Arbor MI 48104

***********@*****.*** Ph: 571-***-****

Chenna Krishna Reddy Varri

***********@*****.*** Ph: 571-***-****

Chenna Krishna Reddy Varri

***********@*****.*** Ph: 571-***-****

Contact this candidate