An Improved YOLOV3 Target Recognition Algorithm Embedded in SENet Structure

Abstract

The invention relates to the field of depth learning, in particular to an improved YOLOV3 target recognition algorithm embedded with a SENet structure, comprising the steps of S100: collecting characteristic information of an object to be recognized and making a data set; the feature information includes image information; S300, taking a part of the data set as a training set and the rest of the data set as a test set; Step S500:embedding SE structure in YOLOV3 algorithm to obtain SE-YOLOV3 algorithm; Step S600: training SE- YOLOV3 on the training set; Step S700: testing SE-YOLOVE3 performanceon the test set. An improved YOLOV3 target recognition algorithm embedded in a SENet structure of the present invention can accurately identify target parts when there are more incomplete parts disturbing in a sample picture, so as to obtain higher precision and recall ratio.

Description

technical field [0001] The invention relates to the field of deep learning, in particular to an improved YOLOV3 target recognition algorithm embedded in a SENet structure. Background technique [0002] Deep learning originated from the study of artificial neural networks, with the purpose of simulating the human brain to acquire and distinguish things. Its characteristic lies in the combination of low-level features to form an abstract high-level, so as to discover the distributed features of data. Convolutional Neural Network (CNN) is a deep learning method that has excellent performance for image processing. [0003] SENet is a convolutional neural network structure proposed by Hu Jie and his team in 2017, and won the ILSVRC2017 classification task champion. The structure consists of three parts: Squeeze, Excitaion, and Reweight, and explicitly builds the interdependence relationship between feature channels. Since it does not change the size of the feature map and does...

AI Technical Summary

Problems solved by technology

Due to the messy position of the parts, there are many incomplete parts in the pictures taken ( Image 6 Edge position), when using YOLOV3 network training, most of the incomplete parts cannot be correctly identified, and there are cases where the background area is misjudged as a part, and the precision rate on the training set is low (only 72.3%), so it needs to be improved YOLOV3 algorithm solves the above problems

Images