Projects

Privacy Resource Reference - anukramani 🔐

Personal knowledge wiki - sangraha 🧑🏽‍💻

–

Sentiment-Predictor-for-Stress-Detection-using-Voice

B.E Project | Datta Meghe College of Engineering | Dec 2019 - May 2020

A Machine Learning Project based on the deep learning model that helps in predicting Emotions which eventually helps in detecting the mental state of the person. Trained the model on the RAVDESS Dataset (24 Actors, 1440 audio files) using the CNN Model with the raw mfcc voice sets. This method helped in improving the overall accuracy from 74% to 93% from the other methods.

Technologies: Python, librosa, pandas, seaborn, glob, keras, numpy, matplotlib, plotty, scipy, tensorflow, sklearn.

Link: https://github.com/sidmulajkar/sentiment-predictor-for-stress-detection

Netflix-Clone-Using-React

Basic Netflix Clone created to understand the concepts of react. The stack that is being used: React(custom hooks, context, portals),Firebase(Firestore/Auth),Compound & styled components. I thought to make everything static (movies, series, shows, …) but then I found this really cool TMDB API through which you can pull data of most popular movies, top rated shows, what’s trending, collections, lists and much more.

Technologies:React(Compound Components), Firebase, Tmdb Api, Git

Link: https://github.com/sidmulajkar/netflix-clone-using-react

Netflix Front Page

Smart-Mirror-with-Voice-Assistant-using-Raspberrypi

People say every second counts and we considered this saying in the best spirit.Customized and integrated a machine learning driven voice assistant model(Alexa), also designed the Mirror front-end with HTML,CSS with basic data fetched in with other api’s called through javascript.The major focus was on to make this very much optimized for the arm-based processer like raspberry. It works! like any other alexa device would work.. with all voice services available.

Technologies: Shell scripting, Html, CSS, C, Javascript, JSON(Api Parsing)

Link: https://github.com/sidmulajkar/smart-mirror-with-voice-assistant