Aerospace Instrument-Making Annotation << Back Neural Network Method for Visual Diagnostics of Aviation Industry Products Bobkov A.V., Nesterov A.S. This paper examines the problem of building a system for automatic detection of aviation products from video images using a detector based on the YOLOv8 neural network. Typical defects and detection methods are discussed, a review of defect detection methods from video images is provided, and the architecture of the YOLO network as the primary algorithm for detecting defects from video images is examined. The detector is tuned using images from the training set, and its performance is assessed. Keywords: automatic defect detection, visual diagnostics, object search and recognition, YOLO network, neural network detector. DOI: 10.25791/aviakosmos.8.2026.1574 Pp. 03-14.