Hello!
My name is Uday, I am a post graduate student at IIT Hyderabad, India.
I am deeply passionate about creating impactful solutions with the help of my AI and Dev skills.
I have 3 years of experience in developing products in Machine learning and Deep Learning in Computer Vision.
Skills
Skilled in Machine Learning, Computer Vision, Deep Learning, Android App Development and Web development.
I have experience in building deep learning models for image classification and detection, multi object tracking, medical data segmentation, anomaly detection, image enhancement, image deduplication, video data classification and key frame identification using state of the art computer vision techniques
Resume
Education
Master of Technology in Computer Science
2019 - 2022
Indian Institute of Technology Hyderabad, India
CGPA: 9.14
Bachelor of Technology in Information Technology
2015 - 2019
Maturi Venkata Subba Rao, Hyderabad, India
Percentage: 82.2%
Professional Experience
Research Assistant
2019 - Present
Indian Institute of Technology Hyderabad, India
- Worked in collaboration with the Japan International Cooperation Agency (JICA) to make Ahmedabad (an Indian city) a smart city (SATREPS project)
- Developed an end to end framework to perform vehicle detection, segmentation, vehicle type estimation, vehicle tracking, vehicle counting and traffic flow analysis.
- Key Learnings: Machine Learning, Deep Learning in Computer Vision, Flask, OpenCV, Tensorfow library
AI Research Intern
DEC 2020 - JUL 2021
Philips Innovation Campus, Bangalore
- Part of Philips Healthcare research team working on impactful healthcare solutions using Artifcial Intelligence.
- Delivered fetal heart view plane classifcation model with 93% accuracy outperforming industry standard results.
- Published a research paper on fetal heart view plane classifcation at OCCUPAI-2021 conference.
- Submitted Invention Disclosures on fetal heart detection and achostic shadows in fetal ultrasound
- Key Learnings: Medical domain profciency, building efcient deep learning architectures on image and video data
Position of Responsibility
Teaching Assistant
Indian Institute of Technology Hyderabad, India
- CS6880 - Multimedia Content Analysis,
- CS6460 - Visual Big Data Analytics
Head of Creative Arts Club
Jan'18 - Mar'19
MVSR FineArts
- Held responsibility for conducting graphic design workshops and building creative event portfolio.
- Managed a team of 20 members with great creative minds to design marketing campaigns.
- Received projects to develop event posters and promotional videos for various event collaborations.
Human Resources
Mar'17 - Apr'17
Kaarmic Educations
- Part of recruiting team for the fellowship programme of Talent Accelerator.
- Conducted written tests, group discussions and individual interviews.
Projects
Master's Thesis - Multi Organ Tumour Segmentation
(Guide: Prof.C.Krishna Mohan)
- Developing a novel bio medical image segmentation architecture using transformer as backbone.
- Integrating spatial and channel level attention to the architecture for 2d and 3d image data.
- Building end to end state of the art multi organ tumour segmentation pipeline with domain invariance.
Mobile Image deduplication
| OPPO R&D
- Built “Image deduplication” pipeline using image hashing, approximate nearest neighbours and image similarity techniques in classic machine learning algorithms.
- Developed a realtime automated to find the duplicate images and remove them from the gallery.
Human Motion Trajectory Prediction
(Guide: Prof.C.Krishna Mohan)
- Predicted human motion trajectory from given observed motion trajectories of the target objects in the scene.
- LSTM is used to predict the future trajectories and GAN to discriminate the generated paths.
Wallet Manager
| Android App Development
- Deployed the "Wallet Manager" application in the play store and has 4.5+ user ratings with 1000+ downloads.
- It enables the user to organise and review their income and expenditure with detail analysis and visualisations.
CollegeSpace
| Social Networking Website
- Designed a social networking website with Frontend: HTML, CSS and JavaScript and Backend: Django.
- It facilitates users to interact with their friends and share their posts with enabled like and comment section.
Contact
Let's get connected..