Domain Adaptive Learning Method

Abstract

The present invention relates to a domain adaptive technology oriented to the field of computer vision. In order to solve the problem of domain offset caused by different distributions of training categories and test categories in the object classification problem, the present invention: a domain adaptive learning method, which is used to represent n samples The training data set, which represents the visual feature space, contains N categories, and uses m test samples from M categories, and the training categories and test categories are disjoint, and each category uses one embedded in the category semantic space. Vector representation, which is attribute space or text description space, p and q represent the dimensions of visual space and category semantic space, the basic principle of cross-modal embedding method is to use training data S to learn a transfer matrix to transfer visual samples from visual space χ to In the category semantic space y, in the test phase, use the transfer matrix W learned in the training phase * Map the test sample x to the category semantic space. The invention is mainly applied to computer vision occasions.

Description

technical field [0001] The present invention relates to a domain adaptive technology oriented to the field of computer vision, specifically, to a domain adaptive learning method. Background technique [0002] In the traditional object classification technology, the category of the test sample must be included in the category that appeared in the training stage. In order to obtain better classification performance, each category often requires a large number of training samples. In reality, samples of some categories are difficult to obtain. For scenarios where there are few or missing training samples in some categories during the training phase, the training model cannot obtain the data distribution of the test category. Using the model learned from the training category to predict the samples of the test category is prone to the problem of domain shift. Take zero-shot classification as an example for introduction. [0003] Zero-shot classification is a scenario where th...

AI Technical Summary

Problems solved by technology

For scenarios where the test category has fewer or missing samples during the training phase, the problem of domain shift is prone to occur.

Images