Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Method for assisting object detection through semantic segmentation

An object detection and semantic segmentation technology, applied in the fields of deep learning and computer vision, can solve the problems of low detection accuracy of small objects, reduced resolution of feature maps, and weakened characteristics of small objects, so as to improve detection performance and strong practicability. and universality, achieving simple effects

Inactive Publication Date: 2019-05-21
TIANJIN UNIV
View PDF4 Cites 36 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In order to extract more robust features, the convolutional neural network needs to perform a pooling operation. The pooling operation will continuously reduce the resolution of the feature map, weaken or lose the features of small objects, and lead to the detection accuracy of small objects. very low

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Method for assisting object detection through semantic segmentation
  • Method for assisting object detection through semantic segmentation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0025] The present invention will be further described below in conjunction with the accompanying drawings.

[0026] Generally speaking, the convolutional network used to extract features in the object detection algorithm can be five modules: Block1, Block2, Block3, Block4, Block5. These five modules are connected in sequence, and the resolution of the feature map is reduced to 1 of the previous one. / 2. Compared with the resolution of the original image, the resolutions of the feature maps of Block1, Block2, Block3, Block4, and Block5 are 1 / 2, 1 / 4, 1 / 8, 1 / 16, and 1 / 32 of the original image in turn. This is achieved through a downsampling operation. Given that the output feature map of Block5 has the highest semantic level, the output feature map of Block5 is usually selected for detection. However, the output feature map of Block5 has low resolution, which is not conducive to small objects and occluded objects.

[0027]The convolutional network that extracts features in th...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention relates to a method for assisting object detection through semantic segmentation. The method comprises the following steps: selecting a trunk network, and setting parameters: semantic segmentation and object detection share the trunk network; designing a spatial information feedback module: fusing the output characteristics of the two modules Block 4' and Block 5' after the semanticsegmentation branch with the output characteristics of the object detection branches Block 4 and Block 5, and feeding the output characteristics of the semantic segmentation branch back to the objectdetection branch to enhance the characteristics of the object detection; design of global attention mechanism module; and designing a semantic segmentation auxiliary object detection structure based on the selected backbone network and the designed spatial information feedback module and global attention mechanism module.

Description

technical field [0001] The invention belongs to the field of deep learning and computer vision, in particular to a method for assisting object detection with semantic segmentation. Background technique [0002] Object detection is a very important area in the field of computer vision. Object detection is the task of locating an object with a rectangular box and identifying the class of the object. Object detection has been widely used in many fields such as video surveillance, autonomous driving, and human-computer interaction. In autonomous driving systems, object detection algorithms can identify preceding vehicles or pedestrians to maintain a safe distance. In video surveillance in traffic scenarios, object detection can assist in the detection of illegal vehicles, etc. [0003] With the improvement of the expression ability of convolutional neural network and the acquisition of big data, the object detection algorithm based on convolutional neural network has made gre...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/62G06N3/08
Inventor 庞彦伟聂晶
Owner TIANJIN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com