Face detection quality scoring method and system

Abstract

The invention discloses a face detection quality scoring method and system. The method comprises the steps: constructing a face detection network, and carrying out the pre-training, so as to enable a model to accurately locate the position of a face; meanwhile, a reward function capable of automatically adjusting rewards and punishment in the training process is provided, and the reward function and the face detection network form an environment generator. A shallow convolutional neural network is utilized to form an intelligent agent to score the face quality. An experience playback strategy and a target Q network algorithm are adopted when the intelligent agent is trained, so that the training speed and the performance of the model can be effectively improved. According to the method, the characteristic that the difference between the faces with different qualities is large is utilized, the deep reinforcement learning thought and the self-adjusting reward and punishment mechanism are combined to score the quality of the faces, the faces with the good quality can be efficiently selected from the video data, and the performance of a face recognition system is improved.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to a face detection quality scoring method and system. Background technique [0002] In recent years, with the rapid development of deep learning technology, face detection technology has made great progress. This is due to the constantly updated advanced neural network architecture and the unremitting efforts of scientific researchers in the theory of face detection. The advancement of face detection technology based on deep learning has also promoted the successful implementation of related application products. Relying on the powerful feature extraction capabilities of deep neural networks and the real-time performance of lightweight neural networks, face detection has been used in campus security and life services. Some good results have been achieved in other fields. [0003] However, for the entire face recognition system, there are still certain problems...

AI Technical Summary

Problems solved by technology

First of all, changes in the external environment are random, unpredictable weather conditions, and the time of day and night will affect the quality of the video; secondly, the facial expressions and postures of people appearing on the screen are also uncertain

Images