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

A framework to enhance the effect of semantic segmentation model based on transfer learning

A technology of semantic segmentation and transfer learning, which is applied in image analysis, image data processing, instruments, etc., to achieve the effect of reducing model speed, improving training, and improving accuracy

Active Publication Date: 2018-12-25
SUN YAT SEN UNIV
View PDF5 Cites 35 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Compared with the first) method, the second) method is much worse in segmentation accuracy, which is also the main shortcoming of the fast semantic segmentation network

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 framework to enhance the effect of semantic segmentation model based on transfer learning
  • A framework to enhance the effect of semantic segmentation model based on transfer learning
  • A framework to enhance the effect of semantic segmentation model based on transfer learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0039] The present invention proposes a new semantic segmentation model framework to improve the accuracy of the fast semantic segmentation network by using the 1) and 2) methods mentioned in the background technology. The scheme of the present invention mainly includes:

[0040] 1) In the first method, the semantic segmentation network with a good segmentation effect but a large and complex model is used as the teacher network, and the semantic segmentation network with fast running speed and poor segmentation effect in the second) method is used as the student network. A new model framework for teacher-student semantic segmentation.

[0041] 2) A pair of complementary 0-order knowledge loss functions and 1-order knowledge loss functions are proposed to transfer the knowledge information of the teacher network to the student network, thereby improving the segmentation accuracy of the student network.

[0042] 3) By using the model in method 1), predict the segmentation label of...

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 framework for improving the effect of semantic segmentation model based on transfer learning, which comprises the following contents: 1) introducing the transfer learning into the semantic segmentation field, so that the fast semantic segmentation network can improve the segmentation effect of the student model through the teacher model; 2) consistency mapping being proposed to measure the contour and texture information of the teacher-student model, and a consistency loss function being constructed to make the fast semantic segmentation better in detail. 3) using a teacher model and a conditional random field (CRF) model to generate auxiliary label for unlabeled data, and adding data to a training set to improve that generalization ability and segmentation effectof the model. The invention improves the accuracy rate of the fast semantic segmentation model without introducing additional model parameter and reducing the model speed.

Description

technical field [0001] The invention belongs to the technical field of semantic segmentation, and in particular relates to a framework for improving the effect of a semantic segmentation model based on transfer learning. Background technique [0002] Image semantic segmentation is a basic and important research in the field of computer vision, which requires the model to identify which semantic category each pixel in the image belongs to. Image semantic segmentation has many applications, such as automatic driving (Automatic driving) and auxiliary robot (Auxiliary robot) and so on. Image semantic segmentation is also the basis of video semantic segmentation. By treating video frames as a single image, the problem can be transformed into semantic segmentation of images, which can be further modeled in the time dimension. [0003] The existing work can be mainly divided into two categories: precision-oriented semantic segmentation and speed-oriented semantic segmentation. Th...

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): G06T7/10
CPCG06T7/10G06T2207/20112G06T2207/20081
Inventor 谢佳锋胡建芳钟逸朱海昇郑伟诗
Owner SUN YAT SEN 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