Image scene classification method and device based on target semantics and attention mechanism
A technology of scene classification and attention, applied in the field of image recognition, to achieve the effect of improving feature expression ability, suppressing useless features, and enhancing useful features
Active Publication Date: 2022-03-25
WUHAN UNIV
View PDF8 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0009] In view of the above defects or improvement needs of the prior art, the present invention proposes an image scene classification method and device based on target semantics and attention mechanism, thereby solving the technical problems existing in the existing scene classification method
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
[0050] In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not constitute a conflict with each other.
[0051] The present invention proposes a natural image scene classification method and device based on target semantic information and attention mechanism, aiming to overcome the difficulties of large intra-category differences and high inter-category similarities of scene pictures, and make full use of the target semantic information of scene pictures Optimize the scene classification results an...
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 discloses an image scene classification method and device based on target semantics and attention mechanism, belonging to the field of image recognition, including: collecting or downloading natural image scene classification data sets; calculating the target probability distribution of each type of scene; The scene classification network of channel attention and spatial attention mechanism is trained on the data set, and then the trained network is used to calculate the category probability of the test picture; for a given test picture, the existing target detection network is used to detect the scene picture Existing targets, and calculate the correction coefficient combined with the results of the previous step; multiply the correction coefficient by the category probability to obtain the final scene classification probability, and output the scene category with the highest probability after sorting. The present invention uses the attention mechanism to extract effective features, suppresses useless features, and combines the target semantic information of the scene, so that the final classification result is more in line with the real scene category, and the accuracy of natural image scene classification is improved.
Description
technical field [0001] The invention belongs to the field of image recognition, and more specifically relates to a natural image scene classification method and device based on target semantic information and attention mechanism. Background technique [0002] With the rapid development of Internet technology and the rapid popularization of mobile devices such as smart phones and SLR cameras, digital images have become a widely used information medium, and the images that can be obtained and accessed are also multiplying. Scene classification is a hot research issue in the field of image recognition, and has a wide range of applications in the fields of retrieval and management of massive images, human-computer interaction, intelligent robots, medical applications, and tourism navigation. The scene is composed of the target, the spatial layout, the background and the relationship between them, and it is a very abstract concept. Scene classification is different from target r...
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): G06V10/764G06V20/00G06V10/774G06K9/62G06N3/04
CPCG06V20/00G06N3/045G06F18/2415G06F18/214
Inventor 陈丽琼邹炼范赐恩王嘉乐程谟凡裘兆炳
Owner WUHAN UNIV