Smit Mahesh Panchal

NLP-Powered Short Video Enrichment System

This research paper presents a smart short video recommendation system powered by Natural Language Processing (NLP). By using techniques like count vectorization and cosine similarity, the system analyzes video hashtags, descriptions, and user IDs to suggest the top five most relevant reels. Built with React, Firebase, and Flask, the solution aims to connect users quickly to content they’ll enjoy, making the recommendation process both efficient and personalized.

Recyclable and Non-Recyclable Waste Recognition using Image Processing with Machine Learning: Recycle Mate

This research paper introduces "Recycle Mate," a machine learning-based system for recognizing recyclable and non-recyclable waste using image processing. By leveraging a Convolutional Neural Network (CNN) built with Keras and integrated via Django, the model efficiently classifies waste items from images into two categories. The system aims to simplify waste sorting for users, improve recycling accuracy, and support smarter, more sustainable waste management practices.

Bhagavad Geeta Based ChatBot

This paper presents a chatbot inspired by the Bhagavad Geeta, designed to answer questions about life, history, and philosophy. Built using the OpenAI API and advanced language models like GPT-2 and GPT-3, the bot interprets user queries and provides thoughtful, context-aware responses based on the teachings of the Geeta. The goal is to make ancient wisdom easily accessible and relevant for today’s users through a simple, interactive chat interface.