Object Detection: Training From Scratch

Kai Zhao,Xin Chen,Yan Zhou

doi:10.1109/access.2020.3018131

Abstract

The development of deep neural networks has driven the development of computer vision. Deep neural networks play an important role in object detection. To improve network performance, before using neural networks for object detection, they are commonly pre-trained on the data set and fine-tuned to their object detection tasks. Pre-training is not always helpful in object detection tasks, so studies have been performed on training neural networks from scratch. By consulting many relevant studies,we performed a systematic analysis of training networks from scratch for object detection. Our article is divided into the following three parts: (i) the reasons for which target detection requires training from scratch, (ii) mainstream networks that can be trained from scratch, and (iii) the criteria for training from scratch. Finally, we summarize some research directions relevant to this topic.

Highlights

C ONVOLUTION Neural Networks (CNNs) play an important role in computer vision and perform well in areas such as object classification, object detection, and semantic segmentation
The features used in traditional object detection algorithms are artificially designed, for example, scale invariant feature transformation [2], histogram of gradient [3], and speeded up robust features [4]
This paper focuses on the application of training from scratch in object detection, most previous reviews have focused on the application of Deep Neural Networks (DNNs) in object detection, and there is no specific discussion on training from scratch

Summary

Introduction

C ONVOLUTION Neural Networks (CNNs) play an important role in computer vision and perform well in areas such as object classification, object detection, and semantic segmentation. Objects in an image are classified into specific categories by pixels. Object detection is used to classify the objects in the image, and to locate the objects. The features used in traditional object detection algorithms are artificially designed, for example, scale invariant feature transformation [2], histogram of gradient [3], and speeded up robust features [4]. These features are used to identify an object, and the object is combined with the corresponding strategy to locate the same

Objectives

Methods

Findings

Conclusion