| My current research is focused on applying Deep Learning to Visual SLAM. This includes improving different components of the Visual SLAM by adding contextual information to the existing pipeline. Previously, I have worked on developing deep learning models for classification, recognition and segmentation problems. I was also involved in design and development of image processing & computer vision algorithms with hands on C-level and multi-threaded optimisation of image processing modules. |
| 2017 - Present | Deep Learning Research Engineer, WhodatTM |
| 2016 - 2017 |
Technical Lead, Samsung R&D India
Media Analytics & Recognition Team |
| 2014 - 2016 |
Lead Engineer, Samsung R&D India Audio, Video & Imaging Solutions Team |
| 2012 - 2014 |
Senior Software Engineer, Samsung R&D India
Multimedia Solutions Team |
| 2010 - 2012 |
Masters in Signal Processing
(5.8/7.0)
Thesis: Complex Network Approach for Analysis of Biomedical signals (ECG, EEG) Indian Institute of Science, Bangalore |
| 2005 - 2009 |
Bachelors in Electronic and Communication Engineering
Sri Jayachamarajendra College of Engineering, Mysore |
| Languages |
C, C++, Matlab, Python |
| Deep Learning Frameworks |
Caffe, TensorFlow |
| OS |
Linux, Windows |
| Miscellaneous Productivity Tools |
Vim, LaTex, Tmux, Visual Studio, Eclipse, Android NDK |
 |
Character Prediction using RNN & LSTM (using numpy library) [code-rnn] [code-lstm] |
 |
Pedestrian Detection using Histogram of Oriented Gradients (HOG) [project page] [code] |
 |
Global Image Descriptor (GIST) [project page] [code] [executable] |
 |
Color Pencil Sketch [project page] [code] [video] [software] |
 |
Basic Computer Vision Toolbox [project page] [video] [software] |
Nrupatunga