Image retrieval model training method, image retrieval method, device, and storage medium
A technology of image retrieval and model training, applied in neural learning methods, biological neural network models, special data processing applications, etc., can solve the problems of large feature differences and low accuracy of similar images, and achieve the effect of improving accuracy
Active Publication Date: 2018-12-28
TENCENT TECH (SHENZHEN) CO LTD
View PDF6 Cites 25 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0005] The embodiment of the present application provides an image retrieval model training method, image retrieval method, equipment, and storage medium, which can solve the problem that the features corresponding to the extracted images of the same type are relatively different when using the image retrieval model in the related art to extract features. The problem of low accuracy in subsequent retrieval of similar images based on feature similarity
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
[0040] In order to make the purpose, technical solution and advantages of the present application clearer, the implementation manners of the present application will be further described in detail below in conjunction with the accompanying drawings.
[0041] For the convenience of understanding, the nouns involved in the embodiments of the present application are described below.
[0042] Convolutional Neural Network (CNN): A deep learning model containing a multi-layer neural network structure, which is widely used in the field of image processing. CNN can be divided into convolution layer, pooling layer and fully connected layer.
[0043] The convolutional layer is a layer used to extract (image) features in a convolutional neural network. It is used to extract low-dimensional features from high-dimensional data, and can include convolution operations and activation operations. Among them, during the convolution operation, the convolution kernel obtained through pre-trainin...
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 retrieval model training method, an image retrieval method, a device and a storage medium, belonging to the image retrieval field. The method comprises the following steps: training an image retrieval model according to an image training set, wherein the image training set comprises a training image with labeling information, and the labeling information is used for indicating a category to which the training image belongs; the training images belonging to the same category being input into the image retrieval model to obtain the image features corresponding tothe training images; determining a feature center point of a class according to image features corresponding to each training image in the same class; according to the image features of the trainingimage, the image retrieval model being trained with the feature center point as the target feature. Embodiments of the present application further train the image retrieval model in a feature clustering manner, so that when feature extraction is performed by using the trained image retrieval model, the feature differences corresponding to the similar images are small, thereby improving the accuracy of subsequent similar image retrieval based on feature similarity.
Description
technical field [0001] The embodiments of the present application relate to the field of image retrieval, and in particular to an image retrieval model training method, image retrieval method, device, and storage medium. Background technique [0002] With the continuous development of deep learning technology, deep learning technology has been widely used in various fields. For example, applying deep learning technology to the field of image retrieval, after a user uploads an image, he can obtain images with similar content to the uploaded image. [0003] In one image retrieval method, developers pre-build an image retrieval model based on deep learning, and use triplet loss (Triplet Loss) as the loss function of the image retrieval model for model training, wherein each triplet contains Anchor examples, Positive examples and Negative examples. When performing image retrieval, after the image to be retrieved is input into the trained image retrieval model, the features of ...
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): G06F17/30G06N3/04G06N3/08
CPCG06N3/08G06N3/045
Inventor 琚震彭湃余宗桥郭晓威
Owner TENCENT TECH (SHENZHEN) CO LTD