Learning apparatus, estimation apparatus, learning method, and non-transitory storage medium

Abstract

The present invention provides a learning apparatus ( 10 ) including: an acquisition unit ( 11 ) that acquires an image; a similarity computation unit ( 12 ) that computes a similarity between the acquired image, and a first image being accumulated in advance and indicating an abnormal state; a registration unit ( 13 ) that registers, as a second image indicating a normal state, the acquired image whose similarity is equal to or less than a first reference value; and a learning unit ( 14 ) that generates an estimation model for discriminating between normal and abnormal by machine learning using the first image and the second image.

What is claimed is: 
     
         1 . A learning apparatus comprising:
 at least one memory configured to store one or more instructions; and   at least one processor configured to execute the one or more instructions to:   acquire an image;   compute a similarity between the acquired image, and a first image being accumulated in advance and indicating an abnormal state;   register, as a second image indicating a normal state, the acquired image whose similarity is equal to or less than a first reference value; and   generate an estimation model for discriminating between normal and abnormal by machine learning using the first image and the second image.   
     
     
         2 . The learning apparatus according to  claim 1 , wherein the processor is further configured to execute the one or more instructions to:
 register, as a third image indicating an abnormal state, the acquired image whose similarity is equal to or more than a second reference value, and   generate the estimation model by machine learning using the first image, the second image, and the third image.   
     
     
         3 . The learning apparatus according to  claim 1 , wherein
 the processor is further configured to execute the one or more instructions to register, as the first image, the acquired image whose similarity is equal to or more than a second reference value.   
     
     
         4 . The learning apparatus according to  claim 1 , wherein
 the processor is further configured to execute the one or more instructions to select a part from among registered images, and generate the estimation model by machine learning using a selected image.   
     
     
         5 . The learning apparatus according to  claim 1 , wherein the processor is further configured to execute the one or more instructions to:
 discriminate a state indicated by the acquired image by using the estimation model,   output the acquired image discriminated to indicate an abnormal state, and accept a correct/incorrect input by a user, and   register, as the first image, the acquired image being input indication of an abnormal state by the correct/incorrect input.   
     
     
         6 . The learning apparatus according to  claim 1 , wherein the processor is further configured to execute the one or more instructions to:
 perform learning of each of a plurality of the estimation models being learned by algorithms different from each other, and   discriminate a state indicated by the acquired image by using each of a plurality of the estimation models, and accumulate a discrimination result of each of a plurality of the estimation models.   
     
     
         7 . The learning apparatus according to  claim 1 , wherein
 the processor is further configured to execute the one or more instructions to acquire an image generated by a surveillance camera.   
     
     
         8 . A learning method comprising,
 by a computer:   acquiring an image;   computing a similarity between the acquired image, and a first image being accumulated in advance and indicating an abnormal state;   registering, as a second image indicating a normal state, the acquired image whose similarity is equal to or less than a first reference value; and   generating an estimation model for discriminating between normal and abnormal by machine learning using the first image and the second image.   
     
     
         9 . A non-transitory storage medium storing a program causing a computer to:
 acquire an image;   compute a similarity between the acquired image, and a first image being accumulated in advance and indicating an abnormal state;   register, as a second image indicating a normal state, the acquired image whose similarity is equal to or less than a first reference value; and   generate an estimation model for discriminating between normal and abnormal by machine learning using the first image and the second image.   
     
     
         10 . An estimation apparatus comprising:
 at least one memory configured to store one or more instructions; and   at least one processor configured to execute the one or more instructions to:   discriminate between normal and abnormal by using an estimation model generated by the learning apparatus according to  claim 1 .