π Dhathri Meda
Applied Machine Learning | Computer Vision | Optimization | Multimodal AI
π M.S. in Applied Machine Learning @ University of Maryland π Building intelligent systems at the intersection of AI, vision, and real-world engineering
π¬ About Me
I work on designing scalable machine learning systems that bridge research and real-world deployment. My experience spans computer vision, optimization, multimodal AI, and LLM-based systems, with a focus on replacing expensive processes with efficient learning-based solutions.
π‘ Research & Work
π± Plant Disease Detection (UMD)
Built an end-to-end deep learning pipeline for multi-species disease classification, improving generalization via augmentation and domain balancing
π Publications Structural Health Monitoring (2025) ASCE Journal of Computing in Civil Engineering (2024) IEEE SNPD (2023) IWSHM Stanford (2023) π οΈ Tech Stack
ML/DL: PyTorch, TensorFlow, XGBoost, ELM Vision: YOLO (v3βv8), SAM, FastSAM, OpenCV AI Systems: LLMs, RAG, Vision-Language Models, Agents Data: Python, Pandas, NumPy, SQL Tools: Docker, Git, Linux, Azure
π Current Focus Multimodal AI (Vision + Language) Autonomous systems & perception Optimization + ML hybrid pipelines Efficient alternatives to simulation-heavy systems π« Connect LinkedIn: https://www.linkedin.com/in/dhathri-meda/ Portfolio: https://portfolio-dhathri-meda.vercel.app/ Google Scholar: https://scholar.google.com/citations?user=svHM7j4AAAAJ&hl=en
