Manoj Surya Kashi

Email : msd.9052@gmail.com

 [Google Scholar  |   Github]

Hello! I am an Engineer at Qualcomm XR group. Prior to this, I worked  with the Advanced Video group of Ittiam Systems, and as a Project Assistant at Visual Information Processing Lab, Indian Institute of Science. I graduated with Bachelors of Technology (B.Tech) degree in Electronics and Communication Engineering with a specialization in Signal Processing from PES University,  India. 

I completed my Bachelor's Thesis in the field of Machine Learning  & Adaptive systems and was advised by Dr. Koshy George. During my undergraduate studies, I was also a part of his Centre for Intelligent Systems lab. Previously, I had a wonderful opportunity to work on Autonomous Robotics and Computer Vision in Microsoft Innovation Lab.   

My interests lie in Computer Vision, Machine Learning, Image and Video signal processing.

Publications

Understanding the perceived quality of video predictions

Nagabhushan Somraj, Manoj Surya Kashi, S. P. Arun, Rajiv Soundararajan

pdf   | Project Website

Meta-Cognitive Neural Networks for Adaptive Identification and Control

Tarun Bhargav Sriram, Manoj Surya Kashi, Rashmi Ugarthod, Koshy George.

31st IEEE Region 10 Conference( TENCON ), 2019

pdf  | slides

An approach to Indoor Mapping using SLAM and Object Detection

Manoj Surya Kashi, Tarun Bhargav Sriram, Rajasekar Mohan.

IEEE Region 10 Symposium( TENSYMP ), 2019

pdf  | slides

Research Experience 

Project Assistant at the Visual Information Processing lab, Indian Institute of Science

Supervisor - Rajiv Soundararajan

I worked on problems related to image and video processing by leveraging machine learning, deep learning, and applied mathematics.

In particular, I am working on problems related to perceptual quality assessment of videos and VR images.



Research Intern, PES Centre for Intelligent Systems, PES University

Supervisor - Dr. Koshy George

I worked on adaptive Identification and control of an system with unknown non-linearity. This involved the learning of the unknown function through neural networks in a sequential manner 


Research Intern,  Microsoft Innovation Lab

Supervisors - Rajasekar Mohan, Dr. Venkatarangan

I worked on developing a robotic platform from scratch that is capable of autonomous navigation.  An article about the project was written by EdexLive.


Teaching Experience

I always had a passion for robotics and it is from the learning experience that I realized that learning robotics can sometimes be circuitous. Thus, when my University initiated a peer learning platform, I was happy and excited to disseminate my knowledge. Through this platform, I designed a 4-week online course titled 'Robotics: A systematic learning approach' and taught about 30 freshman students. I am proud that many of them have gone on to win national-level robotics competitions and are truly excelling.