An improved helmet detection method for YOLOv3 on an unbalanced dataset

TL;DR

This paper enhances YOLOv3's accuracy on unbalanced datasets by using Gaussian fuzzy data augmentation, improving confidence levels and localization without sacrificing speed.

Contribution

It introduces a Gaussian fuzzy data augmentation method to pre-process data, boosting YOLOv3's accuracy and localization performance on unbalanced datasets.

Findings

Confidence level improved by 0.01-0.02

Enhanced image localization performance

No change in recognition speed

Abstract

The YOLOv3 target detection algorithm is widely used in industry due to its high speed and high accuracy, but it has some limitations, such as the accuracy degradation of unbalanced datasets. The YOLOv3 target detection algorithm is based on a Gaussian fuzzy data augmentation approach to pre-process the data set and improve the YOLOv3 target detection algorithm. Through the efficient pre-processing, the confidence level of YOLOv3 is generally improved by 0.01-0.02 without changing the recognition speed of YOLOv3, and the processed images also perform better in image localization due to effective feature fusion, which is more in line with the requirement of recognition speed and accuracy in production.