Projects

• Detection of Insulators from images captured from UAVs by drawing bounding boxes

• Implemented Faster FRCNN architecture

• Compared the test results between the usage of different CNNs (ResNet50 and VGG16)

• Implementation of a multi-node local private Ethereum network using GethClient, which enables blockchain developers to quickly implement projects and perform testing.

• Designed a 'faucet' which deposits a specified amount of Ethers into newly created accounts. This web app had a frontend in Javascript, Backend in NodeJS and was interfaced using Web3 JS.

• Containerisation of the project using Docker

• Developed a complete compiler from scratch using C based on the language rules given in a custom language ERPLAG.

• Independently built all data structures needed from scratch.

• Systemically done in a pipelined method with the following stages:

Lexical Analysis -> Parsing -> AST creation -> Type Extraction -> Semantic Analysis and Type checking ->Code generation.

• Used Deep Learning model for face Recognition with very limited training data.

• Works with different lighting, orientation and additional add-ons(hat,sunglasses etc)

• Pre-trained models are used to extract facial features from single train image to create a database

• Implemented a basic firewall in Prolog.

• The code is segmented into a firewall engine and knowledge base.

• The firewall engine is structured in a way that it is independent of the knowledge base values.

• The System prototype is used to generate a Sine/Triangular/Square waveform of variable Frequencies

• It includes a Complete User Interface designed and simulated using Proteous 8.6.

• The programming was done using Assembly Language (ALP)

• Interactive Sudoku Game designed using C++ in TurboC++.

• Includes a user Interface build using the graphics.h library.

• Helps in both solving problems and generating new puzzles.