html5, CSS3, Bootstrap
javascript (ES6)
MongoDB
Express
React
Node
C++
Python
Data Analysis
Data Visualization
Scikit-Learn
NLTK
CNN
RNN
Keras
Pytorch
OpenCV
Flask API
Recognizes about 80 daily-life objects like person, bike, bags, couch, plants, etc. from images using deep learning model trained on COCO dataset.
Entire frontend is responsive and built with React.
The images are processed on server-side and sent back as API response.
The API is built with Flask served by Gunicorn WSGI server and deployed to Heroku.
Used stack:
objectdetector77.netlify.comSimple PyTorch code for artistic style transfer in images using pretrained VGG-19 convolution network (as feature extractor).
Used stack:
Github RepoSimple OpenCV code to detect faces and eyes in video streams with the help of Haar Cascades. The processed video is saved as an avi video file. The code is meant for processing video files, but with some minor changes, it can be used to detect faces (and eyes) in live video streams and/or images.
Used stack:
Github Repo Minimal full-stack MERN web application with user authentication and interactive dashboard, powered by highly secure MongoDB database, with dual encryption of users' sensitive data like passwords. Entire front end is built on React and is cross-platform and responsive across a variety of devices.
The purpose of this project was to build a PWA that would act as boilerplate for full-stack developers.
Used stack:
valour77.herokuapp.comOpenWeatherMap and Pexels API based Weather PWA. It's built on react with bootstrap. The bakground image is dynamically fetched from pexels, based on the weather conditions of the place searched.
Used stack:
weather77.netlify.com