Projects

Highlight Extractor [Graduate Capstone]

• The web-application suggests best moments in a video taken during a bike ride.
• It used NIMA technical model to get good images within each scene and then used aesthetic model to score them.

Technologies: Flask, OpenCV, Tensorflow, React (in Typescript), PyCharm, WebStorm

Poster Decoder [Course Project]

• Designed and developed an Android image scanner app using opencv.
• The quality of the generated images is comparable to those generated by traditional scanners.

Technologies: Android (in Kotlin), opencv-mobile, Android Studio

Image Manager [Under-Graduate Capstone]

• Designed and developed a swing application for text search on image files.
• The users will assign a tag to a part of an image, that are indexed with Lucene APIs.
• Implemented RPC between master computation desktop and peer mobile applications using Dropbox.

Technologies: Swing (Java), Android, Eclipse, Android Studio

Other Projects

Python

• sudoku-solver: Tries to solve input SuDoKu puzzle using hidden and naked tuples; and backtracking. It uses fixed baseline wherever possible and most constrained variable with backtracking where the first does not solve the problem.

Jekyll

• material-jekyll: Jekyll theme supporting material design with bootstrap 4 for github pages. Also, this page is built using the same theme.

All other repositories can be found at github/sudharkj?repositories.