Research in 2015

Projects

●      Development of a parallel B­spline algorithm for constrained global optimization ​by Prof P S V Nataraj, IIT Bombay, Prof A Zidna, Prof Hoai An Le Thi, Prof Dominique Michel from University of Lorraine, France

●      Development of parallel polynomial global optimization over simplices by Prof P S V Nataraj, IIT Bombay, Prof Priyadarshan Dhabe, VIT Pune, and Prof Deepak Gawali, Vidyavardhini’s College of Engineering & Technology, Mumbai, India and Prof J Garloff, University of Konstanz, Germany

●      Parallelization of 3­D turbulent Navier­Stokes compressible flow solver for the Indian Space Research Organization (ISRO)​ by Gangadhar Vasa, Mandar Gurav, Prof. Sachin Patkar and Prof. Avijit Chatterjee

●      IIT Bombay Autonomous Vehicles Project

○ Vehicles:

SeDriCa​ for Mahindra Rise Prize Driverless Car Challenge

●      Have made it to the working prototype phase and Full Solution Demonstration phase of the challenge. This is scheduled to be held in 15 April and 17 October, 2016 respectively.

Pushpak ​for Intelligent Ground Vehicle Competition

●      This is scheduled to be held from 3 June to 6 June, 2016.

○ Technical Details

■ Uses NVIDIA Jetson TK1 in the data acquisition and mapping module. The mapping module is responsible for gathering data from various sensors such as LiDARs, monocular cameras, stereo cameras and creating a grid map of the surrounding. This map is later used by Path Planner and subsequent modules. Most of the image processing would be done in NVIDIA chipset to get high frequency of input data rate. Some other parallel algorithms such as Bayesian and Particle filters would also be run on the chipset using CUDA libraries. ■ Major Software and Libraries:

1.     Robot Operating System (ROS)

2.     Point Cloud Library (PCL)

3.     OpenCV

4.     Gazebo Simulator

5.     CUDA

● Hardware Details:

1.     RoboteQ (http://www.roboteq.com/index.php​)

2.     2D LiDAR

3.     Monocular cameras

4.     Bumblebee stereo vision camera

           People: ​Arpit Gupta, Rakesh Mistry, Ravi Chalsani, Shubham Jain, Akash

Kishore, Rishabh Choudhary, Ankit Sharma, Rishabh Malviya, Rahul Sharma, Payal Parnami, Abhiroop Rastogi, Tejas Kotwal, Raunaq Bhirangi and Prof. Amarnath Chitta, Prof. S. N. Merchant, Prof. Ramesh Kumar Singh, Prof. Abhishek Gupta,  Prof. V. Kartik from IIT Bombay

●      GPU parallelization of Modified KNN (MKNN) classifier​ by Prof Mrs.

S.G. Lade and Prof. Priyadarshan Dhabe, Vishwakarma Institute of Technology, Pune

●      Accelerating Relational Database query processing on GPUs​ by

Priyadarshan Dhabe, Prof. Mrs. Ashwinee Shingare, VIT Pune

●      Optimization on a very large number of dimensions​ ​ by Dwaracherla

Vikranth, Prof Sachin Patkar, IIT Bombay

●      Speeding up the execution of the 3D unstructured AUSM+UP codes​ ​ by

Gangadhar Vasa, Prof Avijit Chatterjee, IIT Bombay

●      Estimating convex hull of a non­convex function ​​with Saideep Reddy, Prof Vivek Borkar, IIT Bombay

●      Application of Quantum Mechanics (QM) Calculations, Molecular Dynamics (MD) and hybrid QM/MM­MD techniques to study chemical and biochemical reactions​ by Dr Chandan Patel, Prof R B Sunoj, IIT Bombay

             

Undergraduate Projects

●      ​Exact string matching algorithms and speed up their performance using the dynamic parallelism feature of the Kepler GPU​ ​by Sourabh S Shenoy, Prof Neelima Reddy, NMIM Institute of Technology, Nitte, Karnataka, India

 

●      ​Graph analytics using GPU ​​ by Supreetha Shetty, Ananya R, Amoolya Shetty, Prof Neelima Reddy, NMIM Institute of Technology, Nitte, Karnataka, India

 

Master’s Projects

●      ​Stream Data mining using IoT for Intelligent Transport System and prediction and route guidance using GPUs ​b​y Brajesh Deb Ajawani, Prof Surya Durbha at the Centre for Studies in Resources Engineering, IIT Bombay.

●      ​Hyperspectral data processing for land­use/land­cover classification using

Jetson TK1​ ​ by Pankaj Randhe, Prof Surya Durbha at the Centre for Studies in Resources Engineering, IIT Bombay 

●      ​Drowsiness detection of vehicle drivers using GPU​ ​ ​by Sanman Sabane, Prof. Priyadarshan Dhabe at Vishwakarma Institute of Technology, Pune

●      GPGPU Based Acceleration of Tsunami Simulation ​b​y Praveen Singariya and Bhaskar M , Prof Virendra Singh, IIT Bombay

●      ​Homomorphic Encryption schemes using GPU ​ ​by Swapnil Bari, Prof Bernard Menezes, IIT Bombay

●      ​Solving partial differential equations ​ by Avishkar Palegar, Prof Virendra

Singh, IIT Bombay

             

Doctoral projects

●      ​Global Optimization using Bernstein Polynomials and CUDA,​ IIT Bombay. 2010­continuing by Priyadarshan Dhabe

Thesis Supervisor: Prof. P. S. V. Nataraj

●      ​An Interval arithmetic library for GPU computing​ ​, IIT Bombay, 2011­continuing by J. Nayak

Thesis Supervisor: Prof. P. S. V. Nataraj

●      ​Modeling, Simulation, and Control of Boilers using CUDA, ​ ​IIT Bombay, 2011­continuing by S. Unnikrishnan

Thesis Supervisor: Prof. P. S. V. Nataraj

●      ​Parallel QFT,​ ​ IIT Bombay, 2012­ continuing by H. Purohit

Thesis Supervisor: Prof. P. S. V. Nataraj

●      ​Investigation of GPGPU performance for Hyperbolic Partial Differential

Equation using Discontinuous Galerkin Method and Finite Difference Method ​​

­ IIT Bombay, 2013­continuing by Mandar Gurav

Thesis Supervisor: Prof Shivasubramanian Gopalakrishnan

●      ​GPU­enabled fast­feature extraction for flood area classification​ ­​ IIT

Bombay 2013­continuing by Kuldeep Kurte

Thesis Supervisor: Prof. Surya Durbha

●      ​SIFT feature clustering of Earthquake data on GPU using Agglomerative hierarchical clustering approach​ ​ ­ IIT Bombay 2011­continuing by Ujwala Bhangale

Thesis Supervisor: Prof Surya Durbha

●      ​Parallelization of optimization algorithms for solving mixed­integer nonlinear programming problems ​​ ­ IIT Bombay 2012­continuing by Prashant Palkar 

Thesis Supervisor: Prof Ashutosh Mahajan

 

Completed Projects:-

 

●      ​GPU parallelization of fuzzy hyperline segment neural networks ​ ​by Priyadarshan Dhabe, Prof P S V Nataraj, IIT Bombay

●      ​High performance SIFT features clustering of VHR satellite images for disaster management​ ​by Ujwala Bhanage and Prof P S V Nataraj, IIT Bombay

●      ​Synthetic Aperture Radar Image Processing using GPU ​​by Barath Sastha, Prof Sachin Patkar, IIT Bombay

●      ​High Performance Implementation of LDPC Decoding Blocks​ ​by Aishwarya R Goyal, ​Prof. Sachin B. Patkar

●      ​Efficiently parallelize cryptographic applications using NVIDIA GPUs as well as multicores​ ​ by Agrawal Vibhor, Prof. Menezes Bernard, IIT Bombay

 

Publications

●      ​GPU parallelization of fuzzy hyperline segment neural networks for pattern recognition ​ ​by Priyadarshan Dhabe, PhD Scholar, IIT Bombay and Prashant Vyas, CUDA Research Center, VIT Pune at GTC 2015

●      High performance SIFT features clustering of VHR satellite images for disaster management​ ​ by Bhangale Ujwala and Surya Durbha in Proceedings of the Third International Symposium on Women in Computing and Informatics​. ACM, 2015.

●      ​Power Grid Analysis on Parallel Computing Platforms:​ Dash S., Bangera V.,

Kumar V. B. Y., Patkar S. B., Trivedi G at ​MAREW, Microwave and Radio

Electronics Week 2015 25th International Conference Radioelektronika 2015, 14th

Conference on Microwave Techniques COMITE 2015, Pardubice, Czech Republic​, April, 21 ­ 23, 2015

●      ​Framework for Application Mapping over Packet­switched Network of

FPGAs : Case studies​ at ​Second International Workshop on FPGAs for Software

Programmers (FSP 2015, UK​) by Vinay B. Y. Kumar , Pinalkumar Engineer, Mandar Datar, Yatish Turakhia, Saurabh Agarwal, Sanket Diwale and Sachin B. Patkar

●      ​Parameterizable FPGA Framework for Particle Filter Based Object

Tracking in Videtoh​ ​ Ibnyte Ernnagtiinoenearl,  CP.o; nVfeerlemnucreu ognan ​,,  vRo.l;. ,P naotk.,a pr,p S.3.5, ­in4 0​V, L3S­7I  DJaens.i g2n0 15

(VLSID), 2015 28

●      ​GPU implementation of Particle Filter based Object Tracking ​b​y Pinalkumar

J. Engineer, Ayan Mishra, Rajbabu Velmurugan, Sachin Patkar : GPU Technology Conference , GTC 2015, MARCH 17­20, 2015, SILICON VALLEY

●  ​Synthetic Aperture Radar Image Processing by Range Migration Algorithm using Multi­GPUs ​at GPU Technology Conference, Barath Sastha S, Sachin B Patkar and Y. S. Rao : ​GTC 2015, MARCH 17­20, 2015, SILICON VALLEY

●  ​Relaxation Based Circuit Simulation Acceleration over CPU­FPGA ​,​ Vinay

B. Y. Kumar, Kulshreshth Dhiman, Mandar Datar, Akash Pacharne, Harihar

Narayanan and Sachin Patkar : 2016 29th International Conference on VLSI

Design and 2016 15th International Conference on Embedded Systems (VLSID)

●  ​Solving Nonconvex Optimization Problems in Systems and Control: A

Polynomial B­Spline Approach ​i​n ​H. A. Le Thi et al. (eds.), Proc. 3rd Int. Conf

Modelling, Computation and Optimization in Information Systems and

Management Sciences (MCO 2015), pp. 467 – 478, Metz, France, 2015 ​b​y D. D. Gawali, Ahmed Zidna, and P. S. V. Nataraj

●  ​Modified Implicit Bernstein Form and its GPU Parallelization for computing the Bernstein Coefficients of a polynomial​ a​t ​GPU Technology Conference, March 2015, USA ​by P. S. Dhabe and P. S. V. Nataraj

●  ​GPU accelerated B­spline coefficient Computation​ ​at GPU Technology Conference 2015 ​by Nitin Agrawal, Deepak Gawali, P. S. V. Nataraj

●  ​SWIFT ​: ​A Fast Enhanced String Matching Algorithm for Heterogeneous

Architectures ​by Sourabh S Shenoy, Supriya Nayak U, Neelima Reddy at IEEE

High Performance Computing ­Student Research Symposium, 2015, Bengaluru

 

Master's Thesis Work

●  ​Synthetic Aperture Radar Image Processing using GPU ​​, Barath Sastha S

(133070022) under the guidance of Prof. Sachin B. Patkar, Department of

Electrical Engineering, IIT Bombay, June 2015

● ​High Performance Implementation of LDPC Decoding Blocks​ ,​ Aishwarya R

Goyal (10D070018) ​under the guidance of Prof. Sachin B. Patkar, Department of Electrical Engineering, IIT Bombay, June 2015

●​ Circuit Simulation Acceleration on Parallel Computing Platforms​, V​ivek

Bangera (123079015) under the guidance of Prof. Sachin B. Patkar, Department of Electrical Engineering, IIT Bombay, June 2015