Sowmya Sanjana Tuvvara

AI Intern specializing in Python, SQL, and Data Analysis, with a focus on Generative AI. Contributed to the development of an AI-powered research assistant, acquiring practical experience with Large Language Models (LLMs) and research automation techniques.

Minor project: blind assistance system

• Designed a machine learning-based system to assist visually impaired individuals in navigation
• Implemented real-time object detection and distance calculation for safety and obstacle awareness
• Integrated voice feedback (pyttsx3) to provide audio guidance, enabling independent mobility
• Utilized TensorFlow APIs, SSD MobileNet, and Python for model development and testing

Major project: hybrid object detection (YOLOv8 + DETR)

• Developed a hybrid object detection model combining YOLOv8 (real-time detection) with DETR (transformer-based accuracy)
• Implemented a two-stage pipeline with IoU-based fusion and confidence thresholds to refine detection results
• Achieved improved accuracy in complex scenes with overlapping and small objects
• Conducted experiments on the COCO 2017 dataset, balancing speed and precision for real-time applications like surveillance and autonomous systems
• Built using Python, PyTorch, YOLOv8, and DETR frameworks

• Introduction to data science - Infosys Springboard
• Artificial intelligence - Infosys Springboard
• English language for competitive exams – NPTEL
• Public speaking – NPTEL