Machine Learning Intern
During my internship at Druma Technologies, I focused on machine learning and computer vision tasks, contributing to the development of a SaaS-based product. My primary responsibilities included working on object detection using YOLO (You Only Look Once) models, particularly YOLOv3 and YOLOv5. I was involved in customizing data preprocessing pipelines, including data augmentation, resizing, and polygon-to-mask conversion for training on the COCO dataset. I extended PyTorch classes to implement a custom data loader, integrating it into the training pipeline. Additionally, I explored evaluation techniques by calculating mAP (mean Average Precision) and comparing model performance metrics. Furthermore, I contributed to improving the accuracy and efficiency of model training and fine-tuning by experimenting with custom preprocessing logic and augmenting large-scale datasets. I also assisted in running comparative analyses between different YOLO versions and other state-of-the-art models. This hands-on experience strengthened my knowledge of deep learning, image processing techniques, and optimization strategies in machine learning workflows
