Lightweight human face key point detection method and system based on convolutional network and storage medium

Abstract

The invention provides a convolutional network-based lightweight face key point detection method and system, and a storage medium, and the method comprises the steps: employing a multi-task network tocomplete the face detection and face alignment parameter calculation in parallel, and carrying out the alignment of an original inclined face; sending the return face into a light weight key point detection network; detecting face key points, for a multi-face key point detection task,using a pre-training scheme of non-frozen transfer learning, training multiple face key points step by step, and using a parallel face aligning mechanism during training; face rotation return original angle. The method has the beneficial effects that the method is improved according to the characteristics of a face key point detection task, an attention mechanism is introduced to score and select the network output of the convolutional network, and the problem of loss function imbalance of face key point detection is relieved; and a face detection task and a face return parameter calculation task are synchronously trained, so that the efficiency of the overall architecture is improved, and the model complexity is reduced.

Description

technical field [0001] The invention relates to the technical field of face detection, in particular to a method, system and storage medium for lightweight face key point detection based on a convolutional network. Background technique [0002] Face key point detection, also known as face key point positioning, is a branch of biometric technology, which refers to locating the key areas of the face from a given face image, including eyebrows, eyes, nose, and mouth. , facial contours, etc. In the relevant face key point detection algorithm, it is customary to call the set of all key points "shape". The detection methods are roughly divided into three types, namely the traditional method based on the shape model, the method based on the cascaded shape regression, And methods based on deep learning, especially convolutional neural networks. [0003] Convolutional neural networks were originally designed to solve image recognition problems. Early image processing tasks relied ...

AI Technical Summary

Problems solved by technology

[0004] The current mainstream face key point detection algorithms are all based on convolutional networks, but the industrial products of face key point detection need lightweight model support. Although deep learning has greatly promoted the research of face key point detection, the problem But the neural network is becoming more and more complex, and an overly complex model is difficult to become the optimal solution in practice

[0005] To sum up, the main technical bottleneck that really affects the deployment of face key point detection in the industry lies in the lack of targeted and concise models. Although there is no lack of research on lightweight algorithms in academics, most models are aimed at image classification tasks. , it is impossible to propose a solution for the unique problems of face key point detection, such as the imbalance of the loss function of the model and the redundancy of the model architecture design. These technical bottlenecks greatly limit the application of face key point detection in actual products. Application prospects

Images