Article information
2023 , Volume 28, ¹ 6, p.68-80
Borzov S.M., Nezhevenko E.S., Potaturkin O.I.
Investigation of the effectiveness of popular neural network models for detecting objects in the observation field
A brief overview of the main approaches used for solving the problems of detection, determination of coordinates and classification of objects by their images is performed using neural network technologies. The most common YOLO and Faster RCNN methods based on predicting the position of objects are considered using one- and two-stage detectors. At the first stage of the two-stage methods, we identify areas with a high probability containing objects inside themselves by selective search or using a special neural network. At the second stage they are examined by the classifier to determine whether or not they belong to specified classes, as well as to clarify their location and size. One-step methods do not use a separate algorithm to generate regions of interest. Instead, they simultaneously form a certain number of bounding boxes with different parameters in one pass and predict the class of the object. An experimental comparison of the effectiveness of these approaches has been performed. It is noted that single-stage methods allow processing video streams in real time, but they are inferior in accuracy when separating close classes of objects of complex shape.
Keywords: neural network technologies, image processing, object detection, convolutional neural networks, deep learning
Author(s): Borzov Sergey Mikhailovich PhD. Position: Head of Laboratory Office: Institute of Automation and Electrometry SB RAS Address: 630090, Russia, Novosibirsk, Academician Koptyug ave. 1
Phone Office: (383)330-90-33 E-mail: borzov@iae.nsk.su SPIN-code: 7504-7810Nezhevenko Evgeniy Semenovich Dr. Position: Leading research officer Office: Institute of Automation and Electrometry SB RAS Address: 630090, Russia, Novosibirsk, Academician Koptyug ave. 1
Phone Office: (383)330-84-53 E-mail: nedj@iae.nsk.su Potaturkin Oleg Iosifovich Dr. , Professor Position: Head of Research Office: Institute of Automation and Electrometry SB RAS, Novosibirsk State University Address: 630090, Russia, Novosibirsk, Pirogova str., 2
Phone Office: (383)330-40-20 E-mail: potaturkin@iae.nsk.su SPIN-code: 8552-8963 Bibliography link: Borzov S.M., Nezhevenko E.S., Potaturkin O.I. Investigation of the effectiveness of popular neural network models for detecting objects in the observation field // Computational technologies. 2023. V. 28. ¹ 6. P. 68-80
|