Hand Written Digit Recognization
Developed a machine learning model to recognize handwritten digits (0-9) using the MNIST dataset. Implemented a Convolutional Neural Network (CNN) to classify digits from images. Achieved high accuracy by preprocessing data, training the model, and evaluating its performance using metrics like accuracy and precision. Integrated the model into a web application for real-time digit recognition.
Social Media Sentiment Analysis
Performed sentiment analysis on a dataset of tweets/Facebook posts to determine the overall sentiment (positive, negative, neutral) of the posts. Used natural language processing (NLP) techniques to preprocess and analyze the text data. Applied machine learning models such as Naive Bayes or Logistic Regression to classify sentiment and evaluate the model's performance using accuracy and confusion matrix.
Real-Time Explicit Image Filtering Using CNN(BE Project)
Developed an image classification model using Convolutional Neural Networks (CNN) on TensorFlow to identify and categorize distressing and non-distressing images. Created a browser extension that scans web pages, blurs explicit images, and disables hyperlinks, ensuring child-appropriate content. The extension preemptively filters content, providing protection against unsettling visuals and links.
Technologies Used: Python, TensorFlow, CNN, JavaScript, Browser Extension Development.
EchoBot-Virtual talking Friend
 Developed an interactive virtual companion using pyttsx3 for text-to-speech conversion. Users interact via textual
input, and the system responds with synthesized speech output, creating an engaging, conversational experience.
 Technologies Used: Python, pyttsx3, Speech Synthesis, Text-based Interaction.