Semi-supervised hash learning method and device applied to image retrieval
A technology of image retrieval and learning method, applied in the field of neural network, can solve the problems of high human and financial resources, inadvisability, and unsatisfactory deep hash learning effect.
Active Publication Date: 2021-03-12
TSINGHUA UNIV
View PDF3 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, in practical applications, the scale of data is huge (such as hundreds of millions of pictures), and manual labeling of these data for training requires a lot of human and financial resources, which is almost impossible; secondly, data labeling may be the relationship between data pairs, such as social The friend relationship of the network, the relationship defined by the image text description, etc., make it not advisable to directly label the data category
The above problems all lead to the unsatisfactory effect of deep hash learning
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
[0107] Various exemplary embodiments, features, and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. The same reference numbers in the figures indicate functionally identical or similar elements. While various aspects of the embodiments are shown in drawings, the drawings are not necessarily drawn to scale unless specifically indicated.
[0108] The word "exemplary" is used exclusively herein to mean "serving as an example, embodiment, or illustration." Any embodiment described herein as "exemplary" is not necessarily to be construed as superior or better than other embodiments.
[0109] In addition, in order to better illustrate the present disclosure, numerous specific details are given in the following specific implementation manners. It will be understood by those skilled in the art that the present disclosure may be practiced without some of the specific details. In some instances, methods, means, componen...
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 disclosure relates to a semi-supervised hash learning method and device applied to image retrieval. The method includes: inputting sample data into a teacher function for processing, obtaining a hash code output by the teacher function, inputting sample data into a student function for processing, and obtaining a student function output hash code; determine the loss of the student function according to the sample data, the output hash code of the teacher function and the output hash code of the student function; backpropagate the gradient of the loss to the student function to adjust the parameters of the student function and complete the student function One training: determine the adjusted parameters of the teacher function according to the adjusted parameters of the student function and the parameters of the teacher function to be adjusted, and complete one training of the teacher function. The embodiments of the present disclosure can use a small number of labeled sample data and a large number of unlabeled sample data for training. The teacher function and the student function obtained through joint training of the teacher function and the student function can be used to efficiently and accurately obtain the hash code of the data for more efficient retrieval.
Description
technical field [0001] The present disclosure relates to the technical field of neural networks, in particular to a semi-supervised hash learning method and device applied to image retrieval. Background technique [0002] With the development of artificial intelligence and information retrieval technology, there are more and more retrieval requirements for complex data such as images. Taking image retrieval as an example, given an image, we want to find images that are similar at the pixel or semantic level. Due to the complex structure and high dimensionality of images, the efficiency and accuracy of image retrieval has become a difficulty for large-scale image data. [0003] The deep hash learning method is a traditional solution to the retrieval of large-scale complex data (such as images, etc.). Hash learning can learn a specific hash function, and map high-dimensional complex data to short binary hash codes, so that the Hamming distance of hash codes for similar data ...
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): G06F16/51
Inventor 张世枫李建民张钹
Owner TSINGHUA UNIV