Monocular face avatar generation method based on Gaussian point rendering

The invention discloses a monocular face avatar generation method based on Gaussian point rendering. The monocular face avatar generation method comprises the following steps: extracting expression parameters and posture parameters of FLAME from a monocular portrait video; defining an initialization space, a standard space and a deformation space; acquiring Gaussian parameters of the points in the deformation space from the position information of the points in the deformation space and the initialization space; gaussian parameters of the points in the deformation space are input into a renderer, and an image is rendered; performing image loss on the rendered image and the input monocular portrait video, and performing training by minimizing the constraint; adding a point addition and deletion strategy in each training iteration to realize point number increase; and driving the trained specific figure avatar through the driving video. According to the method, an iterative optimization strategy and a Gaussian