An improved helmet detection method for YOLOv3 on an unbalanced dataset

11/09/2020
by   Rui Geng, et al.
0

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.

READ FULL TEXT
POST COMMENT

Comments

There are no comments yet.

Authors

page 2

page 4

08/18/2020

Image Pre-processing on NumtaDB for Bengali Handwritten Digit Recognition

NumtaDB is by far the largest data-set collection for handwritten digits...
09/28/2020

Video Face Recognition System: RetinaFace-mnet-faster and Secondary Search

Face recognition is widely used in the scene. However, different visual ...
01/19/2017

Fast and Efficient Skin Detection for Facial Detection

In this paper, an efficient skin detection system is proposed. The algor...
05/11/2021

Research on Mosaic Image Data Enhancement for Overlapping Ship Targets

The problem of overlapping occlusion in target recognition has been a di...
11/13/2021

Developing a Novel Approach for Periapical Dental Radiographs Segmentation

Image processing techniques has been widely used in dental researches su...
01/03/2022

Line Clipping in E3 with Expected Complexity O(1)

A new line clipping algorithm against convex polyhedron in E3 with an ex...
07/18/2019

A feasibility study of deep neural networks for the recognition of banknotes regarding central bank requirements

This paper contains a feasibility study of deep neural networks for the ...
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.