Local descriptor generation method and device, electronic equipment and computer program product

Abstract

The invention provides a local descriptor generation method and device, electronic equipment and a computer program product, and the method comprises the steps: generating a cross-domain data set which comprises a plurality of image pairs; according to the scale feature map of the cross-domain data set, feature-level domain self-adaptive supervision information is obtained; pixel-level cross-domain consistency supervision information and descriptor loss information are obtained according to the dense descriptors of the feature-level domain self-adaptive supervision information; based on the combination of the feature-level domain adaptive supervision information, the pixel-level cross-domain consistency supervision information and the descriptor loss information, total loss is determined, the total loss is used for supervision training of the network, and a local descriptor is obtained. According to the method, the invariance and robustness of the local descriptor can be enhanced, and the precision of the descriptor on an image matching task is improved.

Description

technical field [0001] The present invention relates to the technical field of computer vision, and in particular, to a method, apparatus, electronic device and computer program product for generating a local descriptor. Background technique [0002] Currently, extracting accurate and efficient local feature descriptors is an essential processing step for various computer vision applications, such as mobile visual search, visual tracking, image matching, image retrieval, and visual localization. While in traditional handcrafted feature extraction methods, SIFT plays a crucial role in computer vision tasks with scale invariance and rotation invariance. Advances in deep learning have resulted in a large gap between the learned local descriptors and classical methods. In recent years, models that jointly learn keypoint detection and descriptors have achieved better performance in image matching and visual localization. Unlike patch-based feature descriptors, these methods ins...

AI Technical Summary

Problems solved by technology

[0004] The present invention provides a local descriptor generation method, device, electronic equipment and computer program products, which are used to solve the defects of low invariance and robustness of descriptors in the prior art, and improve the accuracy of descriptors in image matching tasks. precision

Images