Face Recognition
Programming Lang:- Python
Library:- opencv, keras, sklearn, matplotlib, pandas
Professor Mentor:- Dr N Kishorjit Singh (kishorjit@iiitmanipur.ac.in)
Team:- 1 Member
About:- In this work, Face recognition system is implemented for academic purpose. It is divided into the 2 parts i.e Deep learning and normal similarity function. In the first part Landmark is detected and feature vector is made with the distance ratio. In the second part the a similarity is found with all the stored feature vector. The number of vector who clear there similarity score Thresold is displayed in the ranking order. For, the first part a convolutional neural network was designed which give the position of the 4 landmark. The model was trained on frontal face images each of dimension 96X96. The output of the model was 8 coordinate ratio.
Online Compiler for self Define Language(RDDS)
Programming Lang:- Python
Library:- No external library
Professor Mentor:- Dr Navanath Saharia (nsaharia@iiitmanipur.ac.in)
Phd Student Mentor:- Himangshu Sarma (himangshu@iiitmanipur.ac.in)
Team:- 4 Member
About:- The compiler is build from scratch with module :-
- Lexical analyser
- Parser
- Intermediate Code Generator
- Error handling
- Symbol table
I worked on lexical analyser and parser module of the compiler. Recursive descent parser was implemented according to the grammar.
Question Answering System
Programming Lang:- Python
Library:- NLTK, Pandas, Standford NER(Java)
Professor Mentor:- Dr N Kishorjit Singh (kishorjit@iiitmanipur.ac.in)
Team:- 1 Member
About:- The System was divided into two part that is text processing and question processing. In text processing the given domain corpus had to go thourgh Tokenizer, Parts of Speech taging, Named Entity recognition, Shallow Parsing, and co-reference extraction. In Question Processing the question was first classified i.e basically what type of answer does the question require. A answer is formed according to the question type and depending upon the text matching/similarity score the answer is completed.
Bidirectional Person Counter without MicroController
Link:- Presentation
COMPONENTS:-
- Laser Diodes (2 )
- Photo Diodes (2)
- IC 555 (3)
- Resistors:- 10 Kilo Ohm (8), 1 Kilo Ohm (2), 8 Kilo Ohm (1)
- Transistors (2)
- Capacitors:- 10µF (3), 0.01 µF (1), 1000 µF (1)
- IC:- 7486, 7408, 7404, 4029, 7446
- Seven Segment Displays - 2
Advanced Multi-Switching DTMF Controller With Arduino
About:- In this i asked to design the Arduino code according to the logic for the DTMF.
Code:- Code