Real-world applications of machine learning, from computer vision to cloud-based AI systems
Developed a machine learning model to locate and localize landmarks in images with high accuracy. Implemented an end-to-end pipeline from data preprocessing to model deployment, achieving robust detection across diverse image conditions.
Built a real-time detection system using ResNet50 to identify pests and intruders in agricultural environments. Utilized OpenCV for video processing and implemented alert mechanisms for immediate pest notifications.
Built an ML-based crop recommendation system using Random Forest trained on soil and weather data. Automated ETL pipeline with Apache Airflow, stored data in Snowflake, and deployed on AWS EC2 with a Gemini AI chatbot interface.
Designed a serverless architecture for automatic image caption generation using AWS services. Leveraged Lambda functions, S3 storage, and AI models to create scalable, cost-effective image processing pipeline.