Librería Samer Atenea
Kálamo Books
Librería Elías (Asturias)
Librería Kolima (Madrid)
Librería Proteo (Málaga)
Distributed PySpark and Deep Learning Architectures for Big Data Object Detection delivers a comprehensive engineering treatment of scaling computer vision models across distributed computing clusters using Apache Spark and modern deep learning frameworks. As aerial imagery, cinematic datasets, and high-resolution video archives expand into petabyte-scale volumes, traditional single-node processing pipelines fail to handle computational bottlenecks, memory constraints, and training latency. This monograph establishes the theoretical foundations, distributed architectures, and algorithmic workflows required to construct high-throughput object detection systems.The text covers distributed data processing with PySpark, RDD-based pipelines, Hadoop integration, and MapReduce paradigms alongside state-of-the-art object detection models including YOLOv8 (horizontal and oriented bounding box detection) and Vision Transformers (DETR). It examines benchmark datasets, domain gap adaptation across DOTA and FILMOD datasets, distributed model training, and rigorous performance evaluation metrics such as mean average precision (mAP), precision, recall, and intersection over union. Designed for distributed systems engineers, computer vision researchers, and big data architects, this volume provides rigorous analytical tools to implement scalable, production-grade object detection systems.