My main research interest is primarily in Robot Vision, Machine Learning and their applications that make this world a better place.
This page contains most of the major projects.
Publications
Projects
This page contains most of the major projects.
Publications
- Reconstructing Vehicles from a Single Image: Shape Priors for Road Scene Understanding
J. Krishna Murthy, G.V. Sai Krishna, Falak Chhaya, K Madhava Krishna
Submitted to International Conference of Robotics and Automation (ICRA), 2017.
arXiv preprint arXiv: 1609.09468, 2016. - Monocular Reconstruction of Vehicles: Combining SLAM with Shape Priors
Falak Chhaya, Dinesh Reddy, Sarthak Upadhyay, Visesh Chari, M Zeeshan Zia, K Madhava Krishna
In International Conference of Robotics and Automation (ICRA), May 2016.
Link to Project Page is here - Linear-Chain CRF Based Intersection Recognition
Siddharth Tourani, Falak Chhaya, K Madhava Krishna
In International Conference on Vehicular Electronics and Safety (ICVES), Dec 2014.
Projects
- Motion Segmentation from a freely moving monocular camera (Funded by Renault)
Moving objects are segmented out from a video taken from a monocular camera mounted on a moving vehicle. There is no object detection framework used here and segmentation is done pixel-wise. A pixel is labeled as stationary or moving based on two constraints namely epipolar constraint and flow vector constraint. For more information on these constraints, refer to this paper. CPU and GPU implementations were implemented and tested. This is not a real time system and was done in a team of 3 people under guidance of Prof K Madhava Krishna. - Stereo Visual Odometry (Funded by CAIR-DRDO)
This ongoing project is a part of autonomous navigation in GNSS denied environment. Apart from Stereo Visual Odometry (VO), we also take input from Inertial Measurement Unit (IMU) as well in order to get better localization. Data from these two sensors are fused with Extended Kalman Filter (EKF). This is being done in a team of 4 people under guidance of Prof K Madhava Krishna. - Applications of Machine Learning in Health-Care
This is an ambitious project that I am currently involved in with Remedy Social. This is a project that demands multidisciplinary skills. As of now, I am the only member working on the machine learning aspect of it. - Geometric Approach for contour extraction from Digital Image
This was a course project that I did in a team of 3 people. It is basically implementation of Masters Thesis of P. Tejada (given here). Apart from implementing the thesis, we also improved the performance of the system by introducing an adaptive threshold. Our report can be found here. - Design and Development of a high end Temperature Controller And Discrete device based control circuit for Peak Detector
This was my B.Tech. final semester project carried out at Physical Research Laboratory (PRL), Ahmedabad under guidance of Mr. Shanmugam M. This project had two independent projects. In one of the parts, I had to design a controller for Thermoelectric Cooler which is used to cool Silicon Drift Detector. The other part was to design and implement a control circuit for a peak detector module PH300 as triggers are needed at some specific time for functions like holding the peak and rejecting any other detection while that peak value is being processed through ADC.