An interactive method for processing image annotation tasks based on crowdsourcing platform
A technology of image annotation and interactive method, which is applied in the direction of still image data retrieval, special data processing application, metadata still image retrieval, etc., to increase revenue, improve annotation efficiency and user experience, and optimize upload process and annotation process.
Active Publication Date: 2021-07-23
FUDAN UNIV
View PDF7 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0004] In order to optimize the problems faced in the process of solving image annotation tasks in the crowdsourcing mode mentioned above, and better solve image annotation tasks, the present invention proposes an interactive method for processing image annotation tasks based on a crowdsourcing platform
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0065] In order to more clearly introduce the technical problems and technical solutions to be solved by the present invention, specific implementation examples are introduced below in conjunction with the accompanying drawings.
[0066] The present invention is based on the crowdsourcing platform to process the interactive design framework of the image labeling task as an example figure 1 shown.
[0067] First, install the Ubuntu 14.04.2LTS operating system in VMware Workstation or other virtual machine software, and build a crowdsourcing experimental platform through the crowdsourcing open source framework pybossa (with postgresql as the database).
[0068] attached figure 2 Design process for uploading and annotating interactive methods for image annotation tasks. Based on the theory of user participatory design, the interaction design optimization is carried out in the process of task uploading and task interaction respectively. Optimize the task upload process of the ...
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
Login to View More Abstract
The invention belongs to the technical field of image labeling, and specifically relates to an interactive method for processing image labeling tasks based on a crowdsourcing platform. The present invention first constructs a set of online crowdsourcing work mode, so that the task demander and crowdsourcing members can complete the labeling task on it, and secondly, take the image labeling task as the main research and entry point to design and implement the corresponding interaction method; Specifically, it includes: constructing a classification system for image labeling tasks and building a crowdsourcing platform; establishing design criteria and design processes for interactive methods; designing specific operation steps for interactive methods based on crowdsourcing platform processing image labeling tasks. The method of the present invention improves the labeling efficiency and user experience of crowdsourcing workers, thereby providing image data training sets for scientific research fields such as machine learning more quickly, and using the network crowdsourcing work mode to complete image data labeling work, and can also help disabled people provide new employment and income-enhancing opportunities for disadvantaged groups such as
Description
technical field [0001] The invention belongs to the technical field of image labeling, and in particular relates to an interactive method for solving image labeling tasks on a crowdsourcing platform. Background technique [0002] In the era of big data, analyzing and using existing data in a reasonable way can effectively help us solve many problems. The field of machine learning collects effective labeled data to train and optimize their models and algorithms, so as to better Make decisions and inferences about some things and problems. Then, in order to obtain a better model, not only a good enough algorithm is needed, but also a large amount of labeled data (such as labeled data for image, video, text and other data categories / contents) is required as the basis for training and analysis. The Internet crowdsourcing work mode provides a friendly channel for solving big data labeling work with its fast and low-cost characteristics. At the same time, it also increases more ...
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
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06F16/58G06K9/62
CPCG06F16/5866G06F18/214
Inventor 丁向华严鑫顾宁卢暾
Owner FUDAN UNIV