Method and apparatus for testing liveness
Disclosed is a method and apparatus for testing a liveness, where the liveness test method includes receiving a color image and a photodiode (PD) image of an object from an image sensor comprising a pixel formed of a plurality of PDs, preprocessing the color image and the PD image, and determining a liveness of the object by inputting a result of preprocessing the color image and a result of preprocessing the PD image into a neural network.
1. A liveness test method, comprising:
receiving a color image and a photodiode (PD) image of an object from an image sensor comprising a pixel formed of a plurality of PDs;
preprocessing the color image and the PD image; and
determining a liveness of the object by inputting a result of preprocessing the color image into a first neural network and a result of preprocessing the PD image into a second neural network,
wherein the determining comprises:
calculating a liveness score based on a first output of the first neural network with respect to the color image and a second output of the second neural network with respect to the PD image; and
determining the liveness of the object to be live in response to the liveness score being greater than or equal to a threshold, and
wherein the calculating comprises obtaining the liveness score by adding a result of multiplying the first output by a first weight of the color image to a result of multiplying the second output by a second weight of the PD image, and
wherein the second weight is based on a distance between the image sensor and the object.
2. The liveness test method of claim 1 , wherein the second weight is based on a proportion of a region of the object to the PD image.
3. The liveness test method of claim 1 , wherein the second weight is based on illuminance information of the image sensor.
4. The liveness test method of claim 1 , wherein the determining comprises:
testing a first liveness of the object based on a first output of the first neural network with respect to the result of the inputting of the preprocessed color image;
determining the liveness of the object to be fake, in response to a result of the testing of the first liveness being fake; and
testing a second liveness of the object based on a second output of the second neural network with respect to the result of the inputting of the preprocessed PD image, in response to the result of the testing of the first liveness being live.
5. The liveness test method of claim 1 , wherein the determining comprises:
testing a first liveness of the object based on a first output of the first neural network with respect to the result of the inputting of the preprocessed color image;
determining the liveness of the object to be live, in response to a result of the testing of the first liveness being live; and
testing a second liveness of the object based on a second output of the second neural network with respect to the result of the inputting of the preprocessed PD image, in response to the result of the testing of the first liveness being fake.
6. The liveness test method of claim 1 , wherein the determining comprises:
testing a first liveness of the object based on a first output of the second neural network with respect to the result of the inputting of the preprocessed PD image;
determining the liveness of the object to be fake, in response to a result of the testing of the first liveness being fake; and
testing a second liveness of the object based on a second output of the first neural network with respect to the result of the inputting of the preprocessed color image, in response to the result of the testing of the first liveness being live.
7. The liveness test method of claim 1 , wherein the determining comprises:
testing a first liveness of the object based on a first output of the second neural network with respect to the result of the inputting of the preprocessed PD image;
determining the liveness of the object to be live, in response to a result of the testing of the first liveness being live; and
testing a second liveness of the object based on a second output of the first neural network with respect to the result of the inputting of the preprocessed color image, in response to the result of the testing of the first liveness being fake.
8. A non-transitory computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to perform the liveness test method of claim 1 .
9. A liveness test apparatus, comprising:
a memory configured to store a color image and a photodiode (PD) image of an object received from an image sensor including a pixel formed of a plurality of PDs; and
a processor configured to:
preprocess the color image and the PD image, and
determine a liveness of the object by inputting a result of preprocessing the color image into a first neural network and a result of preprocessing the PD image into a second neural network,
wherein the processor is further configured to:
calculate a liveness score based on a first output of the first neural network with respect to the color image and a second output of the second neural network with respect to the PD image, and
determine the liveness of the object to be live in response to the liveness score being greater than or equal to a threshold,
wherein the processor is further configured to obtain the liveness score by adding a result of multiplying the first output by a first weight of the color image to a result of multiplying the second output by a second weight of the PD image, and
wherein the second weight is based on a proportion of a region of the object to the PD image.
10. The liveness test apparatus of claim 9 , wherein the second weight is based on illuminance information of the PD image.
11. The liveness test apparatus of claim 9 , wherein the processor is further configured to:
test a first liveness of the object based on a first output of the first neural network with respect to the result of the inputting of the preprocessed color image,
determine the liveness of the object to be fake, in response to a result of the testing being fake, and
test a second liveness of the object based on a second output of the second neural network with respect to the result of the inputting of the preprocessed PD image, in response to the result of the testing being live.
12. The liveness test apparatus of claim 9 , wherein the processor is further configured to:
test a first liveness of the object based on a first output of the first neural network with respect to the result of the inputting of the preprocessed color image,
determine the liveness of the object to be live, in response to a result of the testing of the first liveness being live, and
test a second liveness of the object based on a second output of the second neural network with respect to the result of the inputting of the preprocessed PD image, in response to the result of the testing of the first liveness being fake.
13. The liveness test apparatus of claim 9 , wherein the processor is further configured to:
test a first liveness of the object based on a first output of the second neural network with respect to the result of the inputting of the preprocessed PD image,
determine the liveness of the object to be fake, in response to a result of the testing of the first liveness being fake, and
test a second liveness of the object based on a second output of the first neural network with respect to the result of the inputting of the preprocessed color image, in response to the result of the testing being live.
14. The liveness test apparatus of claim 9 , wherein the processor is further configured to:
test a first liveness of the object based on a first output of the second neural network with respect to the result of the inputting of the preprocessed PD image,
determine the liveness of the object to be live, in response to a result of the testing of the first liveness being live, and
test a second liveness of the object based on a second output of the first neural network with respect to the result the inputting of the preprocessed color image, in response to the result of the testing of the first liveness being fake.