Automatic Audio Captioning

Working on a project as part of the DCASE (Detection and Classification of Acoustic Scenes and Events) challenge, which generates descriptive captions for audio clips by analyzing audio features. The system uses deep learning models to transcribe and describe the content of the audio automatically. The goal is to assist accessibility and improve user experiences in multimedia content by providing real-time audio captions, aligning with the task of sound event detection and classification in the DCASE challenge.

Improving Eye Lens Quality Control with AI: Automated Detection of Defective Lenses

This project involves the use of computer vision and deep learning to automate the detection of defective eye lenses during the manufacturing process. The AI model identifies defects such as scratches, cracks, and other imperfections, improving the overall quality control process and reducing human error.

Mental Health Assisting Chatbot

Developed an interactive chatbot using Streamlit to assist students with depression, utilizing NLP for query understanding, personalized suggestions, and a depression assessment tool for sentiment analysis and resource recommendations.

Sound Event Detection using PANN CNN14

Fine-tuned the PANN CNN14 model for sound event detection across 18 classes, integrated it with ROS for real-time classification, and enhanced robustness through data augmentation and feature extraction.

To maintain privacy, certain code details have been omitted.

Smart Attendance Management System

Developed a system to enhance attendance management efficiency, accuracy, and convenience, featuring a user-friendly interface for managing records and generating reports, using HTML, CSS, JavaScript, MS SQL Server, and Flask.

Stress Analysis Project

Conducted stress analysis based on data from 800 participants, identifying key stressors, their impact, and changes in behavior, with a focus on age group differences.

Time Series Analysis for Sales Forecasting at Walmart

Analyzed time series data from 45 Walmart stores using SARIMAX, informing sales forecasting and strategic decisions.

Global Superstore Sales Dashboard

This project involves creating an interactive sales dashboard using Power BI to visualize and analyze sales data from a global superstore.