A sorting method, device, machine-readable medium and equipment

Abstract

The invention discloses a sorting method, which includes: clustering the pictures similar to the retrieved face pictures retrieved from the database to obtain a plurality of clusters; obtaining the retrieved face pictures and each cluster The face similarity score between clusters; according to the face similarity score, sort the retrieved pictures of multiple clusters to get the sorting result. The present invention greatly improves the intelligibility of the retrieval results without affecting the validity of the retrieval results, and prevents users from subjectively rejecting the retrieval results as much as possible.

Description

technical field [0001] The present invention relates to the field of artificial intelligence, in particular to a sorting method, device, machine-readable medium and equipment. Background technique [0002] In face retrieval engine services, precise retrieval is often achieved through high-dimensional face feature vectors extracted by computer deep neural networks. However, the high-dimensional face vectors generated by computer deep neural networks are usually not very intuitive and interpretable. To put it bluntly, the portrait information observed by the computer is richer, more detailed, and more abstract than that observed by the human eye. [0003] The unintuitive and uninterpretable problem of this high-dimensional face feature vector is that the retrieved most similar face images often do not conform to the common sense of human cognition, making it difficult for people to understand and accept. For example, sometimes the closest images retrieved from a face image a...

AI Technical Summary

Problems solved by technology

[0003] The unintuitive and uninterpretable problem of this high-dimensional face feature vector is that the retrieved most similar face images often do not conform to the common sense of human cognition, making it difficult to understand and accept

Images