Research Project, Microsoft Research Lab - India, Bangalore, India
In this work, we address the fundamental problem of structured knowledge extraction from service incidents. We have built SoftNER, a framework for unsupervised knowledge extraction from service incidents.
Research Project, PES Center for Pattern Recognition, PES University, Bangalore, India
Developing a physiotherapy framework employing Microsoft Kinect Sensor to provide an engaging physiotherapy session along with performance analysis to monitor progress in patients.
Bachelor Thesis Project, PES University in collaboration with Rx Digi Health Platform, Bangalore, India
Features extracted from retinal images of premature infants are used to build a model to identify the stage of the disease. This is in collaboration with Rx Digi Health Platform, a start-up based in Bangalore, India.
Class Project (Aug '16 - May '19), PES University, Bangalore, India
Developing a generic web based drawing tool that supports construction, analysis and evaluation of technical diagrams by providing flexibility to developers to plug-in various domains of interest.
Class Project (Aug '18 - Dec '18), PES University, Bangalore, India
A study of various adversarial attacks along with defenses for the attacks for deep networks using ImageNet. A performance analysis of these attacks on standard models was also carried out.
Internship Project (June '18 - July '18), Carnegie Mellon University, Pittsburgh, USA
Analysis of various patterns in the input-output sequences of various obfuscated circuits to define a metric to quantify the level of obfuscation in a circuit using machine learning techniques.
Class Project (Jan '18 - May '18), PES University, Bangalore, India
Identify whether two questions have the same intent or not. GloVe encoding and a network consisting of fully connected ReLU layers followed by sigmoid classifier were employed on the Quora dataset.