Facial animation implementation method, computer device, and storage medium

1. A facial animation implementation method, performed at a computer device having one or more processors and memory storing a plurality of programs to be executed by the one or more processors, the method comprising:

capturing, by the computer device, a facial image of a person;

extracting, by the computer device, facial feature points in the facial image;

comparing, by the computer device, the facial feature points with corresponding standard feature points of a neutral face, to obtain a first deformation coefficient corresponding to a geometrical feature;

extracting, by the computer device, a local region from the facial image according to the facial feature points for processing, to obtain a second deformation coefficient corresponding to an appearance feature by:

determining, by the computer device, the second deformation coefficient corresponding to the local region according to a texture feature, including:

determining, by the computer device, an identification result corresponding to the local region by using a trained classifier according to the texture feature, wherein the classifier is obtained by learning the texture feature in a labeled sample; and

determining, by the computer device, the second deformation coefficient corresponding to the texture feature according to the identification result;

performing, by the computer device, coefficient smoothing processing on the first deformation coefficient and the second deformation coefficient by using a least squares filter method that includes a filter processing window having more than one frame; and

driving, by the computer device, a three-dimensional virtual object by using the first deformation coefficient and the second deformation coefficient that have been smoothed, to perform a corresponding expression represented by the facial image of the person.

2. The method according to claim 1 , wherein the second deformation coefficient comprises a third deformation coefficient and a fourth deformation coefficient; and

the extracting, by the computer device, a local region from the facial image according to the facial feature points for processing, to obtain a second deformation coefficient corresponding to an appearance feature comprises:

extracting, by the computer device, the local region from the facial image according to the facial feature points;

computing, by the computer device, the texture feature corresponding to the local region;

determining, by the computer device, the third deformation coefficient corresponding to the local region according to the texture feature; and

determining, by the computer device, an eyeball center position by using an eyeball tracking algorithm, computing a rotation angle of an eyeball relative to a horizontal visual axis according to the eyeball center position, and determining the fourth deformation coefficient according to the rotation angle, wherein the eyeball center position comprises a position c* of a center of the eyeball corresponding to a maximum sum of dot products of di

c

*

=

arg

⁢

⁢

max

c

⁢

{

1

N

⁢

∑

i

=

1

N

⁢

(

d

i

T

⁢

g

i

)

2

}

,

and gi: where gi is a gradient vector of a position xi, di is a vector from c* to xi, and d i T represents a transpose matrix.

3. The method according to claim 1 , wherein the extracting, by the computer device, the local region from the facial image according to the facial feature points comprises:

converting, by the computer device, the facial image to a standard facial image by using Piecewise Affine Warping according to the facial feature points; and

extracting, by the computer device, the local region from the standard facial image.

4. The method according to claim 1 , wherein the comparing, by the computer device, the facial feature points with standard feature points, to obtain a first deformation coefficient corresponding to a geometrical feature comprises:

computing, by the computer device, three-dimensional coordinates corresponding to the facial feature points; and

comparing, by the computer device, the computed three-dimensional coordinates corresponding to the facial feature points with three-dimensional coordinates corresponding to the standard feature points of the neutral face, to obtain the first deformation coefficient corresponding to the geometrical feature.

5. The method according to claim 4 , wherein the comparing, by the computer device, the obtained three-dimensional coordinates corresponding to the facial feature points with three-dimensional coordinates corresponding to the standard feature points of the neutral face, to obtain the first deformation coefficient corresponding to the geometrical feature comprises:

computing, by the computer device, the three-dimensional coordinates corresponding to the standard feature points of the neutral face;

comparing, by the computer device, the computed three-dimensional coordinates corresponding to the facial feature points with the three-dimensional coordinates, to determine change values corresponding to parts of a face; and

performing processing on the determined change values to obtain the first deformation coefficient.

6. A computer device, comprising one or more processors, memory coupled to the one or more processors and a plurality of programs stored in the memory that, when executed by the one or more processors, cause the computer device to perform a plurality of operations comprising:

capturing, by the computer device, a facial image of a person;

extracting, by the computer device, facial feature points in the facial image;

comparing, by the computer device, the facial feature points with corresponding standard feature points of a neutral face, to obtain a first deformation coefficient corresponding to a geometrical feature;

extracting, by the computer device, a local region from the facial image according to the facial feature points for processing, to obtain a second deformation coefficient corresponding to an appearance feature by:

determining, by the computer device, the second deformation coefficient corresponding to the local region according to a texture feature, including:

determining, by the computer device, an identification result corresponding to the local region by using a trained classifier according to the texture feature, wherein the classifier is obtained by learning the texture feature in a labeled sample; and

determining, by the computer device, the second deformation coefficient corresponding to the texture feature according to the identification result;

extracting, by the computer device, a local region according to the facial feature points for processing, to obtain a second deformation coefficient corresponding to an appearance feature;

performing, by the computer device, coefficient smoothing processing on the first deformation coefficient and the second deformation coefficient by using a least squares filter method that includes a filter processing window having more than one frame; and

driving, by the computer device, a three-dimensional virtual object by using the first deformation coefficient and the second deformation coefficient that have been smoothed, to perform a corresponding expression represented by the facial image of the person.

7. The computer device according to claim 6 , wherein the second deformation coefficient comprises a third deformation coefficient and a fourth deformation coefficient; and

the extracting, by the computer device, a local region according to the facial feature points for processing, to obtain a second deformation coefficient corresponding to an appearance feature comprises:

extracting, by the computer device, the local region from the facial image according to the facial feature points;

computing, by the computer device, the texture feature corresponding to the local region;

determining, by the computer device, the third deformation coefficient corresponding to the local region according to the texture feature; and

determining, by the computer device, an eyeball center position by using an eyeball tracking algorithm, computing a rotation angle of an eyeball relative to a horizontal visual axis according to the eyeball center position, and determining the fourth deformation coefficient according to the rotation angle, wherein the eyeball center position comprises a position c* of a center of the eyeball corresponding to a maximum sum of dot products of di

c

*

=

arg

⁢

⁢

max

c

⁢

{

1

N

⁢

∑

i

=

1

N

⁢

(

d

i

T

⁢

g

i

)

2

}

,

and gi: where gi is a gradient vector of a position xi, di is a vector from c* to xi, and d i T represents a transpose matrix.

8. The computer device according to claim 7 , wherein the extracting, by the computer device, the local region from the facial image according to the facial feature points comprises:

converting, by the computer device, the facial image to a standard facial image by using Piecewise Affine Warping according to the facial feature points; and

extracting, by the computer device, the local region from the standard facial image.

9. The computer device according to claim 6 , wherein the comparing, by the computer device, the facial feature points with standard feature points, to obtain a first deformation coefficient corresponding to a geometrical feature comprises:

computing, by the computer device, three-dimensional coordinates corresponding to the facial feature points; and

comparing, by the computer device, the obtained three-dimensional coordinates corresponding to the facial feature points with three-dimensional coordinates corresponding to the standard feature points of the neutral face, to obtain the first deformation coefficient corresponding to the geometrical feature.

10. The computer device according to claim 9 , wherein the comparing, by the computer device, the obtained three-dimensional coordinates corresponding to the facial feature points with three-dimensional coordinates corresponding to the standard feature points of the neutral face, to obtain the first deformation coefficient corresponding to the geometrical feature comprises:

computing, by the computer device, the three-dimensional coordinates corresponding to the standard feature points of the neutral face;

comparing, by the computer device, the obtained three-dimensional coordinates corresponding to the facial feature points with the three-dimensional coordinates, to determine change values corresponding to parts of a face; and

performing processing on the determined change values to obtain the first deformation coefficient.

11. A non-transitory computer readable storage medium storing a plurality of machine readable instructions in connection with a computer device having one or more processors, wherein the plurality of machine readable instructions, when executed by the one or more processors, cause the computer device to perform a plurality of operations including:

capturing, by the computer device, a facial image of a person;

extracting, by the computer device, facial feature points in the facial image;

comparing, by the computer device, the facial feature points with corresponding standard feature points of a neutral face, to obtain a first deformation coefficient corresponding to a geometrical feature;

extracting, by the computer device, a local region from the facial image according to the facial feature points for processing, to obtain a second deformation coefficient corresponding to an appearance feature by:

determining, by the computer device, the second deformation coefficient corresponding to the local region according to a texture feature, including:

determining, by the computer device, an identification result corresponding to the local region by using a trained classifier according to the texture feature, wherein the classifier is obtained by learning the texture feature in a labeled sample; and

determining, by the computer device, the second deformation coefficient corresponding to the texture feature according to the identification result;

extracting, by the computer device, a local region according to the facial feature points for processing, to obtain a second deformation coefficient corresponding to an appearance feature;

performing, by the computer device, coefficient smoothing processing on the first deformation coefficient and the second deformation coefficient by using a least squares filter method that includes a filter processing window having more than one frame; and

driving, by the computer device, a three-dimensional virtual object by using the first deformation coefficient and the second deformation coefficient that have been smoothed, to perform a corresponding expression represented by the facial image of the person.

12. The non-transitory computer readable storage medium according to claim 11 , wherein the second deformation coefficient comprises a third deformation coefficient and a fourth deformation coefficient; and

the extracting, by the computer device, a local region according to the facial feature points for processing, to obtain a second deformation coefficient corresponding to an appearance feature comprises:

extracting, by the computer device, the local region from the facial image according to the facial feature points;

computing, by the computer device, the texture feature corresponding to the local region;

determining, by the computer device, the third deformation coefficient corresponding to the local region according to the texture feature; and

determining, by the computer device, an eyeball center position by using an eyeball tracking algorithm, computing a rotation angle of an eyeball relative to a horizontal visual state according to the eyeball center position, and determining the fourth deformation coefficient according to the rotation angle, wherein the eyeball center position comprises a position c* of a center of the eyeball corresponding to a maximum sum of dot products of di

c

*

=

arg

⁢

⁢

max

c

⁢

{

1

N

⁢

∑

i

=

1

N

⁢

(

d

i

T

⁢

g

i

)

2

}

,

and gi: where gi is a gradient vector of a position xi, di is a vector from c* to xi, and d i T represents a transpose matrix.

13. The non-transitory computer readable storage medium according to claim 12 , wherein the extracting, by the computer device, the local region from the facial image according to the facial feature points comprises:

converting, by the computer device, the facial image to a standard facial image by using Piecewise Affine Warping according to the facial feature points; and

extracting, by the computer device, the local region from the standard facial image.

14. The non-transitory computer readable storage medium according to claim 11 , wherein the comparing, by the computer device, the facial feature points with standard feature points, to obtain a first deformation coefficient corresponding to a geometrical feature comprises:

computing, by the computer device, three-dimensional coordinates corresponding to the facial feature points of the neutral face; and

comparing, by the computer device, the obtained three-dimensional coordinates corresponding to the facial feature points with three-dimensional coordinates corresponding to the standard feature points of the neutral face, to obtain the first deformation coefficient corresponding to the geometrical feature.