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

Target object property recognition method and device

An attribute recognition and target object technology, applied in the field of image recognition, can solve the problems of poor robustness, low accuracy of pedestrian attribute recognition, lack of data, etc., to achieve the effect of improving accuracy and robustness

Active Publication Date: 2018-02-16
BEIJING KUANGSHI TECH +1
View PDF6 Cites 39 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, during the research process, the inventors found that the process of pedestrian attribute recognition using the above-mentioned conventional method is easily disturbed by factors that are not conducive to attribute recognition, such as changes in human body posture and partial occlusion of the human body, and the robustness is poor; in addition, There are also problems such as lack of data, which will lead to low accuracy of pedestrian attribute recognition

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
  • Target object property recognition method and device
  • Target object property recognition method and device
  • Target object property recognition method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0027] First, refer to figure 1 An example electronic device 100 for implementing the method and device for identifying the attribute of a target object according to the embodiment of the present invention will be described.

[0028] Such as figure 1 As shown, the electronic device 100 includes one or more processors 102, one or more storage devices 104, an input device 106, an output device 108, and an image acquisition device 110, and these components are connected through a bus system 112 and / or other forms of connection mechanisms (not shown) interconnects. It should be noted that figure 1 The components and structure of the electronic device 100 shown are only exemplary, not limiting, and the electronic device may also have other components and structures as required.

[0029] The processor 102 may be a central processing unit (CPU) or other forms of processing units with data processing capabilities and / or instruction execution capabilities, and may control other comp...

Embodiment 2

[0036] see figure 2 The flow chart of a method for attribute recognition of a target object is shown, the method is implemented by a recognition network, the recognition network includes a feature extraction network, and a pedestrian re-identification network, a pedestrian gesture recognition network, and a pedestrian attribute recognition network respectively connected to the feature extraction network network; the parameters of the recognition network are jointly trained by the pedestrian re-identification network, the pedestrian pose recognition network and the pedestrian attribute recognition network based on the shared feature extraction network; it can be understood that the parameters of the recognition network include the parameters of the feature extraction network, pedestrian re-identification The parameters of the recognition network, the parameters of the pedestrian pose recognition network, and the parameters of the pedestrian attribute recognition network. The p...

Embodiment 3

[0075] For the attribute recognition method of the target object provided in the second embodiment, the embodiment of the present invention provides a device for attribute recognition of the target object, which is applied to the recognition network, and the recognition network includes a feature extraction network, and a feature extraction network and a feature extraction network respectively. The network-connected pedestrian re-identification network, pedestrian gesture recognition network and pedestrian attribute recognition network; the parameters of the recognition network are jointly trained by the pedestrian re-identification network, pedestrian gesture recognition network and pedestrian attribute recognition network based on the shared feature extraction network. For a detailed explanation about the identification network, refer to the foregoing embodiments.

[0076] see Figure 4 A structural block diagram of an attribute recognition device for a target object as show...

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 provides a target object property recognition method and device and relates to the field of image recognition technology. The method is executed by the adoption of a recognition network.The recognition network comprises a feature extraction network, a pedestrian re-recognition network, a pedestrian posture recognition network and a pedestrian property recognition network, wherein the pedestrian re-recognition network, the pedestrian posture recognition network and the pedestrian property recognition network are connected with the feature extraction network, and parameters of therecognition network are obtained through joint training by the pedestrian re-recognition network, the pedestrian posture recognition network and the pedestrian property recognition network based on the mode of sharing the feature extraction network. The method comprises the steps that the feature extraction network extracts feature information of a target object; and the pedestrian property recognition network determines properties of the target object based on the feature information. Through the target object property recognition method and device, the pedestrian property recognition network can acquire richer and more comprehensive feature information by means of sharing the feature extraction network with the pedestrian re-recognition network and the pedestrian posture recognition network, and therefore the accuracy and robustness of property recognition are effectively improved.

Description

technical field [0001] The present invention relates to the technical field of image recognition, in particular to an attribute recognition method and device of a target object. Background technique [0002] Pedestrian attribute recognition technology can identify the attribute information of pedestrians in the picture; among them, the attribute information of pedestrians can include gender, age, color of clothes, whether they have backpacks, etc. Pedestrian attribute recognition technology usually adopts the method of: using SVM (support vector machine, support vector machine) or convolutional neural network to train several two-category classifiers to judge whether the current target has a certain attribute, or to determine which attributes need to be judged in total Finally, train a multi-class classifier to judge these attributes. However, during the research process, the inventors found that the process of pedestrian attribute recognition using the above-mentioned conv...

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
IPC IPC(8): G06K9/00G06K9/46G06K9/62G06N3/08
CPCG06N3/084G06V40/20G06V10/40G06F18/24
Inventor 肖琦琦张弛
Owner BEIJING KUANGSHI TECH
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