|
dbavikad at asu dot edu I'm a second-year PhD student advised by Dr.Paulo Shakarian at ASU. Prior to starting my PhD, I completed Integrated Masters of Technology in Information Communication Technology at the School of Engineering, Central University of Karnataka. I worked as a research assitant at Lab1055,IITH as a part of Prof.Vineeth N Balasubramanian's research group. Also, I worked on a research project in Medical Imaging under Prof.Jayanthi Sivaswamy at IIITH. I completed my Bachelors in Electronics and Communications Enginnering(ECE), ranking 1 in the school of engineering. I pursued my research in Speech Emotion Recognition, advised by Prof.Dr.Anil Kumar at IIIT Hyderabad. I was fortunate enough to intern at Bharat Sanchar Nigam Limited (BSNL: a prestigious telecommunications service in India), Advanced Training Institute for Electronics and Process Instrumentation (ATI-EPI: of Gov. of India), and Develearn technologies as a data science researcher. I'm the founder and editor-in-chief for the IEEE SB magazine- Tarang. I also like science fiction, travelling, origami, making art, badminton, chess, constructive debates. |
|
|
I'm interested in Computer Vision, Signal and Speech Processing using AI, Video Analysis and Understanding, Vision+Speech+Language, Deep Learning, Limited Supervision, Internet of Things, and their medical applications. I am passionate in Research and Development, I would like to take up a fundamental or applicative task in Research field with commitment, honesty and hardworking to meet forth coming challenges in the Science and Technology by absorbing day-to-day innovations. The desire to learn more and more about everything is the driving force behind a lot that I do. |
|
|
|
|
Keywords: pose recognition, explainable AI, computer vision. Lab 1055, IITH in colaboration with Variance AI An application that takes a real time input of a person doing a complex excercise like yoga or of physiotherapy or even martial arts and recognizes the pose. It is also able to correct the pose if its done wrong. We also contribute two novel first of its kind datasets to the vision community. |
.PNG)
|
Keywords: face recognition, Deep CNN, computer vision. Develearn Technologies, Mumbai A smart multi-purpose end-to-end approach for an automated system that successfully marks attendance of the staff, maintains a log of details of the staff,gives various features for the staff to check upcoming holidays, apply for a leave etc. The developed system has a novel architecture that uses MTCNN-Facenet framework integrated with a front end that doesn't require a large database that gives an improved false positive rate (18% better), true positive rate (14% greater) than literature. |
|
Keywords: RNN, CNN, LSTM, MLP. B.Tech Thesis, International Institute of Information Technology, Hyderabad(IIITH)+ Central University of Karnataka Developed an integrated complex state-of-the-art system that recognizes emotions from a live stream, and is embedded with a front end that includes a web application with advanced real-time 3D visuals and multiple features, and a backend with the deep learning architecture that is wrapped with multiple different models: RNN, CNN, LSTM, MLP, that were built from scratch, trained on a GPU and an interfacing layer between the frontend and backend, all which in entirety works in a single protocol flow which is also designed from scratch. The accuracy against the test database has increased up to 89.74%, the emotions trained are more that range in number of nine and also is speaker independent. |
|
Keywords: Unet, Brain MRIs. IIITH
|
|
~Everything that you see around here is prone to change drastically. I'm not sure when everything will be pulled together, but I can assure it's somewhere around this month.~ |