Interactive face image attribute marking device based on MFC

Abstract

The invention provides an interactive face image attribute marking device based on MFC (Microsoft Foundation Classes), and the method comprises the steps: step a, face detection step: a face image is loaded, faces in the image are detected through a face detector, and face positions are saved; step b, default label generation step: default labels of attributes of the faces which are detected in the step a and displayed on a display interface in sequence are generated and displayed on the interface; step c, label correction step: after the generation of default labels of attributes, and when the displayed attributes of the faces does not match with the default labels is discovered, matched label items are selected from pull-down menus of unmatched labels; step d, missed face marking step: when a face is missed to detect, the face needs to mark is selected, and the matched labels are selected according to the step c; and step e, marking information saving step: face image data with marked attributes is saved. According to the invention, the efficiency and precision for face attribute marking task can be improved.

Description

technical field [0001] The invention relates to the technical field of face recognition, in particular to an interactive face image attribute labeling method based on MFC (Microsoft Foundation Classes, Microsoft Foundation Class Library). Background technique [0002] Face attribute recognition, including gender, age, whether to wear glasses, whether it is blurry, etc., is increasingly used in face multi-task recognition, and the annotation of these attributes comes from manual annotation. The development of face technology is becoming more and more mature. The demand for attribute resources and description text is massive. At present, face attribute annotation is basically manual annotation, which is complex, slow, and inefficient. Therefore, there is a need for a face attribute labeling method with a friendly interface, simple operation, and high efficiency, which can provide a large amount of accurate description text for face multi-task learning and speed up the developm...

AI Technical Summary

Problems solved by technology

[0002] Face attribute recognition, including gender, age, whether to wear glasses, whether it is blurry, etc., is increasingly used in face multi-task recognition, and the annotation of these attributes comes from manual annotation. The development of face technology is becoming more and more mature. The demand for attribute resources and description text is massive. At present, face attribute labeling is basically manual labeling, which is complex, slow, and inefficient.

Images