What I've done and what I'm doing.

textify-VQA is a Vision-Question-Answering (VQA) project focused on answering product-related questions using both visual and textual information from product images.

The Career Crafters Interview Analysis System is an AI-powered platform designed to simulate interviews, assess both verbal and non-verbal communication skills, and provide detailed feedback to improve performance.

Developed a banking chatbot using Ollama to classify intents and perform tasks such as retrieving account balance and transferring funds.

UrbanGraphNet is a custom Graph Neural Network architecture that fuses GAT and Transformer components to model and generate urban road networks.

unique graph-based cycle detection method for identifying craters.

PanScope is a smart document parsing system that extracts structured information from Indian PAN cards.

Using the Multi-Agent Deep Deterministic Policy Gradient (MADDPG) algorithm, the system improves performance and enhances resource management.

A Simple chat app where you can chat and msg anyone

A web-based Campus Delivery System designed to streamline food, grocery, and medicine delivery across the IIT Bhilai campus.

MultiTaskNLP is a deep learning project designed to solve multiple natural language processing tasks — Named Entity Recognition (NER), Sentiment Classification, and Emotion Detection — simultaneously using a shared encoder architecture.

QuoteMind is a semantic quote retrieval application that lets you find meaningful quotes based on natural language queries.

A concept-based explainability framework applied to large multimodal models (LMMs) using the VQAv2-small dataset.

Meta-learning approaches (Black-box and MAML) evaluated on the Omniglot dataset for few-shot image classification.

This project demonstrates a zero-shot text classification approach on the AG-News dataset using pretrained language models.

This is the 1st version of Labeling application powered with AI for Doctors and Healthcare workers.