My name is
Sharik Ali Ansari
My interests are machine learning and artificial intelligence. I am inspired by Andrew ng. He says "AI is a super power which you can use to solve big problems world face today."
My name is
Detection of neurological disorders
using eye based features
Earlier it is belived that you can't detect neurological disorders using visual methods. We introduced a fully atomated vision based method for detection of three neurological disorders.
Buckling based Railway Accidents Prevention
using Deep Learning
The problem of buckling of railway track is least studied but is responsible for many accidents. We tried to provide a solution for this.
Computer Vision Syndrome
(cvs) prevention
Eye related problems that arise due to our eye's constant exposure to different type of screens. Anyone who stares computer, phone or TV screen for a period of more than 3hrs is under risk of aquiring CVS.
Railway Accidents Prevention
(All types of train accidents)
Railway accidents occur due to many reasons. Collisions of human or animals with train or the bad weather including landslides, flooding of railway track or displacement of track. All of these reasons were considered and a solution is presented.
Audio Based gun type detection
Rifle vs handguns
The audio from a handgun and rifle differ very much. Experts often recognise the difference very easily. Here we train a model to detect the type of gun that fired the bullet.
Automated Riots Detection
using CCTv footage
Different parts of the world are facing public Linching and mass riots. Due to already established security cam infrastructure they can potentially be used to detect and report such incidences to authorities. In this project we implemented this solution.
RNN Implemetation
From scratch with just numpy.
CNN Implemetation
From scratch with just numpy.
ANN Implemetation
From scratch with just numpy.
Reinfrocement Q Learning
Using gym library for environment.
CUDA Python
Massive parallelism of program.
LSTM Models
Showing use of different LSTM layers.
Chatbot
Q and A Using LSTM.
Language Converter
Ger to Eng using LSTM.
RCNN
Using CNN and openCV.
Human Activity Recognition
Using 3D-CNN.
Bar and QR code
Generator and Reader.