Sneha Maddela

Computer Vision & AI-based Image Processing Currently working on research and experimental development in computer vision and deep learning for industrial image analysis. My work focuses on investigating how different structured image representations influence visual feature quality and downstream analysis. I am implementing and comparing different approaches for image segmentation, region representation, feature aggregation, and deep feature extraction. The internship involves designing systematic experi

As a Working Student, I contributed to the development of AI-driven perception systems for traffic and railway environments. I created structured annotation guidelines and object class lists, labeled large-scale datasets using CVAT, and trained optimized YOLO models on multi-GPU clusters. I streamlined dataset management with DVC, experiment tracking with MLflow, and automated model training workflows via SSH/tmux, ensuring efficient, reproducible, and high-performance results.

Collaborated on sensor fusion algorithms for object detection, ensuring compliance with ISO 26262 safety standards.

Strengthened project coordination and team collaboration skills while consistently adapting to new challenges in a fast-paced environment.