Supercharge Your Innovation With Domain-Expert AI Agents!

A Semi-Supervised Salient Object Detection Method

An object detection and semi-supervised technology, applied in the field of convolutional neural networks, can solve the problem of heavy workload in training mode, and achieve the effect of reducing workload and saving time consumption

Active Publication Date: 2020-09-29
广西荷福智能科技有限公司
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to overcome the deficiencies of the prior art, provide a semi-supervised salient object detection method, and solve the problem that the prior art adopts a full-supervised training mode with a large workload

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
  • A Semi-Supervised Salient Object Detection Method
  • A Semi-Supervised Salient Object Detection Method
  • A Semi-Supervised Salient Object Detection Method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0029] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings: a semi-supervised salient object detection method, based on the improved RPN network model of faster rcnn, such as figure 1 As shown, the RPN network model includes an object detection module and a salient prediction module; the input terminals of the object detection module and the salient prediction module share a shared convolutional layer; the shared convolutional layer includes multiple convolutional layers.

[0030] The salient prediction module includes a first convolutional layer, a sigmoid layer and three transposed convolutional layers. Among them, the convolution kernel size of the first convolution layer is 1*1, the stride is 1, the pad is [0, 0, 0, 0], the number of convolution kernels is 1, and the output of this layer is a feature Spectrum, which is consistent with the output size of the previous layer (ie, the last convoluti...

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 discloses a semi-supervised salient object detection method, which is based on an improved faster rcnn RPN network model. The method includes: S1: performing initial segmentation on the object frame in the picture, as the initial ground truth of the object; S2: setting up the network; S3: Train the network; S4: Predict salient objects on the training pictures; S5: Perform superpixel segmentation and superpixel-level smoothing on the training pictures, and perform binarization operations; S6: Use the binarized objects obtained in step S5 The foreground spectrum is used as ground truth; S7: Repeat steps S3~S6. The invention proposes a joint network including a salient prediction module and an object detection module, and fuses the output of the salient prediction module with the intermediate feature layer of the classification module for joint optimization. This network structure effectively utilizes the contour information of objects, and assists the network to detect salient objects more accurately.

Description

technical field [0001] The invention relates to the field of convolutional neural networks for salient object detection, in particular to a semi-supervised salient object detection method. Background technique [0002] Humans have the ability to perceive objects of interest from pictures. In computer vision, such objects of interest are called salient objects. The ability of computers to imitate humans to detect salient objects from pictures is called salient object detection. Salient object detection has become a popular field because it can assist other image processing tasks and has a wide range of applications, including object detection, object segmentation, scene understanding, image classification and retrieval, etc. [0003] In the field of computer vision, there are many salient object detection methods. These methods can be divided into two categories: one is unsupervised salient object detection methods, and the other is supervised salient object detection meth...

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 Patents(China)
IPC IPC(8): G06K9/62G06N3/04
CPCG06N3/045G06F18/2155
Inventor 崔静谭凯王嘉欣
Owner 广西荷福智能科技有限公司
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com