Estimation device, estimation method, and storage medium

Abstract

An estimation device according to one aspect of the present disclosure includes: at least one memory storing a set of instructions; and at least one processor configured to execute the set of instructions to: generate a plurality of extraction regions by adding a perturbation to an extraction region of a partial image determined based on positions of feature points extracted from a face image; estimate a plurality of directions of at least one of a face and a line of sight and a reliability of each of the plurality of directions based on a plurality of partial images in the plurality of extraction regions of the face image; and calculate an integrated direction obtained by integrating the plurality of directions based on the estimated reliability.

1 . An estimation device comprising:
 at least one memory storing a set of instructions; and   at least one processor configured to execute the set of instructions to: 
 extracting a first partial image from a face image, the first partial image including a region of the face image determined based on a position of a feature point in the face image; 
 evaluating an image quality of the face image; 
 determining parameters according to the image quality of the face image; 
 generating a second partial image by adding perturbation to the region of the face image, wherein magnitude of the perturbation added to the region of the face image is determined based on the parameter; 
 estimating a plurality of directions and reliabilities of the plurality of directions from the first partial image and the second partial image, the plurality of directions being a plurality of face directions or a plurality of directions of lines of sight; and 
 calculating an integrated direction based on the reliabilities, the integrated direction being a direction into which the plurality of directions is integrated. 
   
     
     
         2 . The estimation device according to  claim 1 , wherein 
 the perturbation is according to a parameter indicating either one of an intensity or an amount of noise added to the region of the face image.   
     
     
         3 . The estimation device according to  claim 1 , wherein 
 the perturbation is according to a parameter indicating an intensity of smoothing of the face image.   
     
     
         4 . The estimation device according to  claim 1 , wherein 
 the image quality of the face image is evaluated based on an amount of noise contained in the face image.   
     
     
         5 . The estimation device according to  claim 1 , wherein 
 the image quality of the face image is evaluated based on a magnitude of a contrast.   
     
     
         6 . The estimation device according to  claim 1 , wherein 
 the perturbation is determined in such a way that the magnitude of the perturbation increases as the image quality of the face image decreases.   
     
     
         7 . An estimation method comprising:
 extracting a first partial image from a face image, the first partial image including a region of the face image determined based on a position of a feature point in the face image;   evaluating an image quality of the face image;   determining parameters according to the image quality of the face image;   generating a second partial image by adding perturbation to the region of the face image, wherein magnitude of the perturbation added to the region of the face image is determined based on the parameter;   estimating a plurality of directions and reliabilities of the plurality of directions from the first partial image and the second partial image, the plurality of directions being a plurality of face directions or a plurality of directions of lines of sight; and   calculating an integrated direction based on the reliabilities, the integrated direction being a direction into which the plurality of directions is integrated.   
     
     
         8 . The estimation method according to  claim 7 , wherein 
 the perturbation is according to a parameter indicating either one of an intensity or an amount of noise added to the region of the face image.   
     
     
         9 . The estimation method according to  claim 7 , wherein 
 the perturbation is according to a parameter indicating an intensity of smoothing of the face image.   
     
     
         10 . The estimation method according to  claim 7 , wherein 
 the image quality of the face image is evaluated based on an amount of noise contained in the face image.   
     
     
         11 . The estimation method according to  claim 7 , wherein 
 the image quality of the face image is evaluated based on a magnitude of a contrast.   
     
     
         12 . The estimation method according to  claim 7 , wherein 
 the perturbation is determined in such a way that the magnitude of the perturbation increases as the image quality of the face image decreases.   
     
     
         13 . A non-transitory program recording medium that records a program causing a computer to execute processing of:
 extracting a first partial image from a face image, the first partial image including a region of the face image determined based on a position of a feature point in the face image;   evaluating an image quality of the face image;   determining parameters according to the image quality of the face image;   generating a second partial image by adding perturbation to the region of the face image, wherein magnitude of the perturbation added to the region of the face image is determined based on the parameter;   estimating a plurality of directions and reliabilities of the plurality of directions from the first partial image and the second partial image, the plurality of directions being a plurality of face directions or a plurality of directions of lines of sight; and   calculating an integrated direction based on the reliabilities, the integrated direction being a direction into which the plurality of directions is integrated.   
     
     
         14 . The non-transitory program recording medium according to  claim 13 , wherein 
 the perturbation is according to a parameter indicating either one of an intensity or an amount of noise added to the region of the face image.   
     
     
         15 . The non-transitory program recording medium according to  claim 13 , wherein 
 the perturbation is according to a parameter indicating an intensity of smoothing of the face image.   
     
     
         16 . The non-transitory program recording medium according to  claim 13 , wherein 
 the image quality of the face image is evaluated based on an amount of noise contained in the face image.   
     
     
         17 . The non-transitory program recording medium according to  claim 13 , wherein 
 the image quality of the face image is evaluated based on a magnitude of a contrast.   
     
     
         18 . The non-transitory program recording medium according to  claim 13 , wherein 
 the perturbation is determined in such a way that the magnitude of the perturbation increases as the image quality of the face image decreases.