Method and apparatus for model optimization
An optimization method and model technology, applied in the computer field, can solve problems such as different content
Active Publication Date: 2019-01-04
BEIJING DAJIA INTERNET INFORMATION TECH CO LTD
View PDF14 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0003] Although the current image description generation model can generate the content description of the image according to the image features of the image, there are still cases where the content description of the generated image is different from the content expressed by the image itself. Therefore, it is often necessary to optimize the image description generation model
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
[0054] In order to make the above objects, features and advantages of the present invention more comprehensible, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.
[0055] refer to figure 1 , which shows a flow chart of the steps of an embodiment of a model optimization method of the present invention, which may specifically include the following steps:
[0056]In step S101, the annotation content descriptions of multiple images are acquired;
[0057] In the embodiment of the present invention, for any image among multiple images, technicians can watch the content of the image in advance, then summarize the content description of the image according to the content of the image, and use it as the marked content description of the image , and then store the annotated content description for that image. Therefore, in this step, the stored annotation content description of the image can be obta...
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 embodiment of the invention provides a model optimization method and a device. The method comprise: acquiring annotation content descriptions of a plurality of images; determining a content topicincluded in each of the annotated content descriptions; determining a target content topic having a frequency of occurrence lower than a preset frequency among content topics included in the pluralityof annotated content descriptions, respectively; acquiring image features of an image corresponding to a target annotation content description including a target content topic; the preset image description generation model being optimized several times by using the target annotation content description and the image feature. As such, when acquiring content descriptions of other images using the resulting preset image description generation model, if the content theme in the content expressed by the other image itself includes a target content theme that occurs very infrequently, the content description of the other image is not different from the content expressed by the other image itself, so that the accuracy of the final obtained preset image description generation model can be improved.
Description
technical field [0001] The invention relates to the field of computer technology, in particular to a model optimization method and device. Background technique [0002] Recently, deep learning has been widely used in related fields such as video, image, speech, and natural language processing. Neural network is an important branch of deep learning. Due to its strong fitting ability and end-to-end global optimization ability, the accuracy of image content description generation tasks is greatly improved after applying the neural network model. [0003] Although the current image description generation model can generate the content description of the image according to the image features of the image, there are still cases where the content description of the generated image is different from the content expressed by the image itself. Therefore, it is often necessary to optimize the image description generation model. Contents of the invention [0004] In order to solve 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
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/62
CPCG06F18/29
Inventor 李岩李涛
Owner BEIJING DAJIA INTERNET INFORMATION TECH CO LTD