Face expression detection method, apparatus, device, system and medium

Abstract

The invention discloses a facial expression detection method, an apparatus, a device, a system and a computer-readable storage medium. The method comprises the following steps: receiving a video picture of each frame transmitted by a head-mounted camera; performing face detection on the video picture according to a preset face detection model; When a human face is detected in the video picture, the coordinates of the key points of the human face are obtained according to a key point detection model established in advance; Detecting facial expression according to the key point coordinates and the static key point coordinates corresponding to the face stored in advance; The invention simply and efficiently realizes the facial expression detection process, improves the operation efficiency and realizes the comprehensive utilization of the operation resources.

Description

Technical field [0001] The present invention relates to the field of computer technology, in particular to a method, device, equipment, system and computer-readable storage medium for detecting facial expressions. Background technique [0002] In recent years, with the development of computer graphics, computer animation technology to synthesize human facial expression animation is one of the focuses of current animation research. It enables animators to directly drive the image model with the actor’s expression, which is influenced by animation. Producers pay more attention. Facial expression capture can detect and record the performer’s expression in real time, and convert it into a digital "abstract expression" so that the animation software can "gave" it to the model so that the model has the same expression as the performer and generates the final expression animation sequence . [0003] The existing more mature expression capture and detection technology is the marker-point...

AI Technical Summary

Problems solved by technology

Although this technology can achieve a high degree of synchronization of body and expression data, due to too many facial detail markers, the large field of view capture suitable for body capture tends to fail to track dense marker points on the face, resulting in a large number of jumping points, and the post-processing is very complicated. , has become the biggest obstacle to practical application, and this solution needs to paste markers on the user's face, which affects the user's actions and also reduces the user's experience; Once the face is out of the capture range of the camera or the head posture changes to a large extent, the relevant detection for the face will become useless, resulting in data interruption or inaccuracy in the results, and also increases the equipment cost. Burden of detecting invalid images, speeding up equipment wear-and-tear

Images