Robin Chacko

- Accelerating AI applications on heterogeneous Embedded devices (FPGA, AI engine) (C++/Python) - Provide Training and conference talks for companies/individuals on accelerating Machine learning application using Xilinx Vitis AI Tools - Backend (Python) Development - Research and investigate the relation between cryptocurrency market returns and investors' sentiments using Machine learning practices.

Learning and Incorporating shape prior information to a Fully Convolutional Network (FCN) Objective:The goal of this project is to encode additional geometry information of the object that has been segmented by a deep learning model to increase the overall accuracy of the network. In particular, we investigate the potential of latent space, i.e. the compressed representation of an image in an encoder-decoder FCN to integrate the new information.

Instance segmentation of persons from RGB images to produce precise silhouettes. Used an existing implementation of state of the art image segmentation method MaskRCNN and adapted its configuration to segments person from the images. Ensure the model's generalization by combing various datasets along with different image augmentation approaches. Finally, the model was trained on different backbone architecture like resnet100/50 with various hyperparameter settings.

Evaluation of Real time 2D Human Pose Detection using an existing Deep Learning model which predict Human pose heatmaps .Real time predictions are integrated with the implementation of Bounding Box tracker to make key point detection faster.

Supervise Lab session for C++ Programming language . Examination and grading the work done by students.

Mathematical Modelling and Control Design of an Autonomous Quadcopter. Primary work was concerned with Simulation of Quadcopter Dynamics and PID controller design for the Six degrees of motion.

Implementation of Manually operating Air suspension system on commercial passenger vehicles. Evaluation of the technical drawing of the recently developed Air suspension system and assistance with selection of suitable particulars for the integration of the new system to the existing Coaches and Buses in the Plant

Deep Learning,Robotics, Control Systems, Power Electronics, Sensorics, Actorics, Machine Dynamics, C++ Programming,Embedded Systems and Control, Electrical and Electronics Engineering, Digital Control, Mechatronics Systems, Software Engineering

Mathematics,Solid Mechanics, Fluid Mechanics, Thermodynamics, C++ Programming, CAD/CAM,Aerospace and Propulsion, Refrigeration, Heat and Mass transfer, Control System, Electrical Engineering.