Tejashree Baddula

Multi-Modal Disaster Response AI System

Collaborated on an AI-powered disaster management project that integrates machine learning models to analyze emergency call audio, social media SOS messages, and satellite imagery. The system uses Whisper for speech-to-text transcription, BERT-based NLP for emergency classification, and a ResNet-based model for satellite image damage assessment. A custom Fusion Engine combines outputs from these modules to generate a unified severity score, helping prioritize rescue operations and enable faster emergency response through real-time alerts and an interactive dashboard.

An Explainable AI Framework for Continuous Fetal Health Monitoring and Early Risk Prediction

Collaborated on a research project proposing an Explainable AI framework for fetal health classification using Cardiotocography (CTG) data. The framework integrates machine learning models such as Random Forest, SVM, and XGBoost with Optuna-based hyperparameter optimization and SHAP/LIME explanations to improve prediction transparency. It also proposes continuous monitoring and early risk assessment to support clinical decision-making.