my website link:

Blogger.com link:

Created a tool that identifies the defective products (ball bearings) in industries using convolutional neural network trained model through the camera vision instead of eye vision.
Train the model over 1000 product train images which is containing both defective and non defective product images in equal percentage.(50% each)
It also prepares the data sheet about the production details and it sends the details to the manager.
Build a client facing API using Django frame work.

For my final year project assignment, I build an underwater object visualizing and detecting API.
This could be useful to find the object under the turbid and dark side of the water.
It is also useful for detecting humans who were trapped inside the water.
created an API that makes the project easy access using Django.
To scan the underwater object, I used VL53LOX which is a small range distance finding sensor.(long-range sensors(Lidar) are available to apparate in deep water level)
By collecting the millions of height data points, the API gives the 3 Dimensional View of underwater objects.
For detecting object type, the 3d view converted into a 2D image then passing into the CNN model which is already a trained model with over 100 scanned underwater object 2D images.

To avoid my Boredom in this lockdown i used to create visitors welcoming bot for my home using face recognition python library, which is audio based bot that recognize the pretrained faces of our visiters and it used to welcome them with their names.
If any untrained faces detected, it consider the name as new visitor and it captures that face to ask the face name later to train that face into the recognition model.
Next time the same faces detected, it welcome them with the name,as what we had denoted already.
This record the each visitors arrival time and departure time also using MySQL database.
This system integrated into an API using Django for easy access.
This system can also be useful for office attendance system with some improvements.

Created a tool that predicts whether the given input mail is spam or ham.
To achieve this Natural Language Processing and support vector machine method used.
This model has trained over 6000 ham and spam emails with an accuracy score of 97.5%.
project 5: Breast Cancer Prediction (Benign vs Melignent):
Created an API that predicts whether the cancer is Benign or Melignent.
trained over 600 datas of benign and melignent cancer attributes. undependent variables are such as radius, texture, perimeter, area, smoothness, etc.
target variables are Benign as 1 and Melignent as 0.

Created the coding contest submission details page along with the code mirror page.
Submission details page consists of User Id, submission Time, Problem Name, Result, Memory, And Compile time.
It also consists of sorting and filtering access.
USED LANGUAGES:
Html, CSS, Java Script, Vue.js, gRPC basics

Done the Assignment of ‘pirple.com’ ‘front end web developement’ course, which is ‘create the replica of any one popular website’. I choose GitHub website.
I created the replica of Home page of the Github website as it is.

For practice, Created an dynamic travel website which is used to book the tourism bus for famous turist places.
The Frame Work done by Django with Html and css. jinja templating language also used.
For Data handelling and storage purpose MySQL database is used.
This website containing user registeration process.