Image Hash code training model algorithm and classification learning method based on binary weight
Active Publication Date: 2018-09-18
CHENGDU KOALA URAN TECH CO LTD
View PDF6 Cites 3 Cited by
Summary
Abstract
Description
Claims
Application Information
AI Technical Summary
This helps you quickly interpret patents by identifying the three key elements:
Problems solved by technology
Method used
Benefits of technology
Problems solved by technology
[0010] The purpose of the present invention is to solve the problems of existing image classification algorithms in large-scale image data sets, such as excessive memory usage, high computational overhead, and poor classification results, and provide a method that can be used in various imag
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
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
Smart Image
Examples
Experimental program
Comparison scheme
Effect test
Example Embodiment
[0066] Example 1
[0067] A hash code image training model algorithm based on binary weights. The algorithm includes the following steps:
[0068] Step 1.1. Select the exponential loss function as the loss function used, and the objective equation formula is:
[0069]
[0070]
[0071] Suppose the generated image binary code is b i Is the original training data set Middle x i The corresponding r-bit binary code, suppose a linear hash equation is:
[0072] b=sgn(P T x)
[0073] Here P is the image hash transpose matrix; T is the transpose symbol; d is the dimension of the image x; r is the length of the hash code;
[0074] The binary code of the classifier is w.
[0075] Step 1.2. Perform unified learning of the binary code of the classifier and training image features obtained in step 1.1, update the training image feature hash code and the binary code of the classifier, and optimize the target equation of the loss function selected in step 1.1, and get optimized After the image hash ...
Example Embodiment
[0105] Example 2
[0106] On the basis of the first embodiment, different loss functions are selected, and the steps of a hash code image training model algorithm based on binary weights are as follows:
[0107] Step 1.1. Choose a simple linear loss function as the loss function used. The objective equation formula is:
[0108]
[0109]
[0110] Suppose the generated image binary code is b i Is the original training data set Middle x i The corresponding r-bit binary code, suppose a linear hash equation is:
[0111] b=sgn(P T x)
[0112] Here P is the image hash transpose matrix; T is the transpose symbol; d is the dimension of the image x; r is the length of the hash code;
[0113] The binary code of the classifier is w.
[0114] Step 1.2. Perform unified learning of the binary code of the classifier and training image features obtained in step 1.1, update the training image feature hash code and the binary code of the classifier, and optimize the target equation of the loss function s...
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 a Hash code image training model based on binary weight, and a model algorithm comprises the steps: selecting a loss function, determining a target equation, and performing binary coding of a classifier and training image features; performing unified learning of a binary code, updating the binary code, and optimizing the loss function; and deducing the Hash code training model. The invention also discloses a classification learning method employing the Hash code image training model based on binary weight, and the method comprises the steps: obtaining a Hash code of a to-be-searched image through the Hash code training model based on binary weight, and solving Hamming distances between the Hash code and a classifier binary code; searching in the minimum Hamming distance from the Hamming distances, and obtaining the classifier corresponding to the minimum Hamming distance, wherein the classifier is the category to which the to-be-searched image belongs. The method can be used for the image classification for various types of images in high-latitude scenes, improves the performance of an algorithm in a large-scale data set, is precise, efficient and quick, andis small in consumption of the memory.
Description
technical field [0001] The invention belongs to the field of image classification methods, and in particular relates to a binary weight-based image hash code training model and a classification learning method. Background technique [0002] In recent years, due to the explosive growth of the number of digital images and the substantial improvement of image quality, the problem of large-scale visual recognition has attracted great research enthusiasm from academia and industry. The classification problem of images with tens of thousands of categories usually uses conventional classifiers for heavy calculations, such as k-nearest neighbors or k-NN, and support vector machines or SVM. In multi-class image recognition problems, a large number of classifiers produces huge computational and memory overheads, and a large number of classifiers will lead to a surge in complexity during the model training and deployment stages. Imagine that there are C categories, and each category h...
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
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.