Method and apparatus with fake fingerprint detection

1. A processor-implemented method, comprising:

obtaining an enrollment fingerprint embedding vector corresponding to an enrollment fingerprint image; and

generating a virtual enrollment fingerprint embedding vector, wherein the virtual enrollment fingerprint embedding vector has an environmental characteristic different from an environmental characteristic of the enrollment fingerprint image, and has a structural characteristic of the enrollment fingerprint image.

2. The method of claim 1 , further comprising:

receiving an input fingerprint image; and

determining whether an input fingerprint included in the input fingerprint image is a fake fingerprint based on a fake fingerprint embedding vector that is provided in advance, the enrollment fingerprint embedding vector, and the virtual enrollment fingerprint embedding vector.

3. The method of claim 1 , wherein the generating of the virtual enrollment fingerprint embedding vector comprises:

obtaining a virtual enrollment fingerprint image by inputting the enrollment fingerprint image to an artificial neural network (ANN); and

generating the virtual enrollment fingerprint embedding vector corresponding to the virtual enrollment fingerprint image.

4. The method of claim 1 , wherein the generating of the virtual enrollment fingerprint embedding vector comprises generating a plurality of virtual enrollment fingerprint embedding vector sets having different environmental characteristics by inputting the enrollment fingerprint image to a plurality of artificial neural networks (ANNs).

5. The method of claim 1 , further comprising:

generating a virtual fake fingerprint embedding vector by inputting the enrollment fingerprint image to an artificial neural network (ANN).

6. The method of claim 5 , further comprising:

receiving an input fingerprint image; and

determining whether an input fingerprint included in the input fingerprint image is a fake fingerprint, based on a fake fingerprint embedding vector that is provided in advance, the enrollment fingerprint embedding vector, the virtual enrollment fingerprint embedding vector, and the virtual fake fingerprint embedding vector.

7. A non-transitory computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to perform the method of claim 1 .

8. A processor-implemented method, method comprising:

receiving an input fingerprint image; and

determining whether an input fingerprint included in the input fingerprint image is a fake fingerprint, based on a fake fingerprint embedding vector, an enrollment fingerprint embedding vector, and a virtual enrollment fingerprint embedding vector that are provided in advance,

wherein the enrollment fingerprint embedding vector is obtained based on an enrollment fingerprint image, and the virtual enrollment fingerprint embedding vector is generated by inputting the enrollment fingerprint image to an artificial neural network (ANN).

9. The method of claim 8 , wherein the virtual enrollment fingerprint embedding vector is generated to have an environmental characteristic different from an environmental characteristic of the enrollment fingerprint image and maintain a structural characteristic of the enrollment fingerprint image.

10. The method of claim 8 , further comprising:

obtaining an input fingerprint embedding vector corresponding to the input fingerprint image,

wherein the determining of whether the input fingerprint is the fake fingerprint comprises:

determining a confidence value of the input fingerprint embedding vector based on the fake fingerprint embedding vector, the enrollment fingerprint embedding vector and the virtual enrollment fingerprint embedding vector; and

determining, based on the confidence value, whether the input fingerprint is the fake fingerprint.

11. The method of claim 8 , further comprising:

performing user authentication based on a result of the determining of whether the input fingerprint is the fake fingerprint; and

determining whether to provide a user access to one or more features or operations of an apparatus, based on a result of the user authentication.

12. A non-transitory computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to perform the method of claim 8 .

13. An apparatus, comprising:

one or more processors configured to:

obtain an enrollment fingerprint embedding vector corresponding to an enrollment fingerprint image; and

generate a virtual enrollment fingerprint embedding vector, wherein the virtual enrollment fingerprint embedding vector has an environmental characteristic different from an environmental characteristic of the enrollment fingerprint image, and has a structural characteristic of the enrollment fingerprint image.

14. The apparatus of claim 13 , further comprising:

a sensor configured to receive an input fingerprint image,

wherein the one or more processors are further configured to determine whether an input fingerprint included in the input fingerprint image is a fake fingerprint, based on a fake fingerprint embedding vector that is provided in advance, the enrollment fingerprint embedding vector, and the virtual enrollment fingerprint embedding vector.

15. The apparatus of claim 13 , wherein the one or more processors are further configured to:

obtain a virtual enrollment fingerprint image by inputting the enrollment fingerprint image to an artificial neural network (ANN); and

generate the virtual enrollment fingerprint embedding vector corresponding to the virtual enrollment fingerprint image.

16. The apparatus of claim 13 , wherein the one or more processors are further configured to generate a plurality of virtual enrollment fingerprint embedding vector sets having different environmental characteristics by inputting the enrollment fingerprint image to a plurality of artificial neural networks (ANNs).

17. The apparatus of claim 13 , wherein the one or more processors are further configured to generate a virtual fake fingerprint embedding vector by inputting the enrollment fingerprint image to an artificial neural network (ANN).

18. The apparatus of claim 17 , further comprising:

a sensor configured to receive an input fingerprint image,

wherein the one or more processors are further configured to determine whether an input fingerprint included in the input fingerprint image is a fake fingerprint based on a fake fingerprint embedding vector that is provided in advance, the enrollment fingerprint embedding vector, the virtual enrollment fingerprint embedding vector, and the virtual fake fingerprint embedding vector.

19. An apparatus, the apparatus comprising:

a sensor configured to receive an input fingerprint image; and

one or more processors configured to determine whether an input fingerprint included in the input fingerprint image is a fake fingerprint, based on a fake fingerprint embedding vector, an enrollment fingerprint embedding vector, and a virtual enrollment fingerprint embedding vector that are provided in advance,

wherein the enrollment fingerprint embedding vector is obtained based on an enrollment fingerprint image, and the virtual enrollment fingerprint embedding vector is generated by inputting the enrollment fingerprint image to an artificial neural network (ANN).

20. The apparatus of claim 19 , wherein the virtual enrollment fingerprint embedding vector is generated to have an environmental characteristic different from an environmental characteristic of the enrollment fingerprint image and maintain a structural characteristic of the enrollment fingerprint image.

21. The apparatus of claim 19 , wherein the one or more processors are further configured to:

obtain an input fingerprint embedding vector corresponding to the input fingerprint image;

determine a confidence value of the input fingerprint embedding vector based on the fake fingerprint embedding vector, the enrollment fingerprint embedding vector, and the virtual enrollment fingerprint embedding vector; and

determine, based on the confidence value, whether the input fingerprint input fingerprint image is the fake fingerprint.

22. A processor-implemented method, comprising:

obtaining an enrollment fingerprint embedding vector corresponding to an enrollment fingerprint image;

generating a first virtual enrollment fingerprint embedding vector having a structural characteristic of the enrollment fingerprint image and an environmental characteristic corresponding to a dry fingerprint; and

generating a second virtual enrollment fingerprint embedding vector having the structural characteristic of the enrollment fingerprint image and an environmental characteristic corresponding to a wet fingerprint.

23. The method of claim 22 , further comprising:

receiving an input fingerprint image; and

determining whether an input fingerprint included in the input fingerprint image is a fake fingerprint based on a fake fingerprint embedding vector stored in a database, the enrollment fingerprint embedding vector, the first virtual enrollment fingerprint embedding vector, and the second virtual enrollment fingerprint embedding vector.

24. The method of claim 22 , wherein the generating of the first virtual enrollment fingerprint embedding vector comprises generating the first virtual enrollment fingerprint embedding vector by inputting the enrollment fingerprint image to a first artificial neural network (ANN), and

wherein the generating of the second virtual enrollment fingerprint embedding vector comprises generating the second virtual enrollment fingerprint embedding vector by inputting the enrollment fingerprint image to a second ANN.

25. The method of claim 24 , further comprising:

generating a virtual fake fingerprint embedding vector by inputting the enrollment fingerprint image to a third ANN;

receiving an input fingerprint image; and

determining whether an input fingerprint included in the input fingerprint image is a fake fingerprint, based on a fake fingerprint embedding vector stored in a database, the enrollment fingerprint embedding vector, the first virtual enrollment fingerprint embedding vector, the second virtual enrollment fingerprint embedding vector, and the virtual fake fingerprint embedding vector.