ASD, a neurological condition, impairs social development. Traditional diagnosis is costly and time-consuming. Recent research uses facial images to detect ASD. This paper uses pre-trained deep learning models classify children's faces, achieving up to 87% accuracy using MobileNet.
This work presents a novel approach to generate accurate disparity maps for rover navigation on celestial bodies with moon-like surfaces. It includes image rectification, disparity map generation, refinement, and occlusion handling. The method outperforms previous approaches in experimental data.
As a part of the Global Infrastructure Operations Database team, we work on handling issues with database and automation.
Worked on the project titled, "Ring Monitors" which focussed on providing an end to end framework for monitoring test rings.
I have worked on extracting and organizing data from programming sites like GeeksforGeeks and Leetcode under the guidance of Dr. Chandan K. Reddy
My team and I have worked on disparity estimation from stereo images for lunar landers. It is a project under the Laboratory for Electro-Optic Systems, peenya unit of ISRO. I was the team leader for the project. The work has been published in Springer.
I was the Teaching Assistant for the course EC253: Principles of Digital Signal Processing, for a batch of over 300. The course was anchored by Dr. B. Niranjana Krupa.
I was the Teaching Assistant for the course EC327: Artificial Neural Networks. The course was anchored by Mrs. Swetha R.
I have worked on “Generating stylized captions for images using deep learning” under the guidance of Ms. P. Ramadevi. Features had been extracted using a pre-trained convolutional neural network. My work included changing the layers of the recurrent neural network using Long Short Term Memory(LSTM) and Gated Recurrent Units(GRU) to generate the captions. The insights gained during the internship are unparalleled and have played a pivotal role in my career.