Generating apparatus, generating method, and non-transitory computer readable storage medium
A generating apparatus according to an embodiment of the present application includes an obtaining unit and a generating unit. The obtaining unit obtains intermediate images in an intermediate layer of a neural network that recognizes the quantity of objects contained in an image. The generating unit generates a combination image by combining together the intermediate images obtained by the obtaining unit. For example, the generating unit generates the combination image by combining together the intermediate images to which a weight has been applied on the basis of a change in the recognition of the quantity of objects corresponding to a process applied to each of the intermediate images.
1. A generating apparatus comprising:
a processor programmed to:
obtain intermediate images in an intermediate layer of a neural network that recognizes a quantity of objects contained in an image; and
generate a combination image by combining together the obtained intermediate images to which a weight has been applied based on a change in recognition of the quantity of objects corresponding to a process applied to each of the intermediate images.
2. The generating apparatus according to claim 1 , wherein the processor is programmed to:
apply a process to the image based on the generated combination image.
3. The generating apparatus according to claim 2 , wherein the processor is programmed to:
remove a part of the image as a processed image based on the generated combination image.
4. The generating apparatus according to claim 3 , wherein the processor is programmed to:
remove the part of the image as the processed image by using a corrected image that is obtained by changing an aspect ratio of the combination image based on an aspect ratio of the image.
5. The generating apparatus according to claim 4 , wherein the processor is programmed to:
remove the processed image from the image based on a feature value in the corrected image.
6. The generating apparatus according to claim 5 , wherein the processor is programmed to:
remove the processed image from the image based on a score calculated from the feature value in the corrected image.
7. A generating apparatus comprising:
a processor programmed to:
obtain intermediate images in an intermediate layer of a neural network that recognizes a quantity of objects contained in an image;
generate a combination image by combining together the obtained intermediate images; and
remove a part of the image as a processed image based on (i) a corrected image that is obtained by changing an aspect ratio of the generated combination image based on an aspect ratio of the image, (ii) a score calculated from a predetermined region in the corrected image, and (iii) a feature value included in the predetermined region in the corrected image.
8. The generating apparatus according to claim 7 , wherein the processor is programmed to:
remove the processed image from the image based on the predetermined region formed by using a predetermined aspect ratio.
9. The generating apparatus according to claim 1 , wherein processor is programmed to:
obtain the intermediate images in the intermediate layer of the neural network that performs a convolution process and a pooling process.
10. A generating method being implemented by a computer and comprising:
an obtaining step of obtaining intermediate images in an intermediate layer of a neural network that recognizes a quantity of objects contained in an image; and
a generating step of generating a combination image by combining together the obtained intermediate images to which a weight has been applied based on a change in recognition of the quantity of objects corresponding to a process applied to each of the intermediate images.
11. A non-transitory computer readable storage medium having stored therein a generating computer program that causes a computer to execute:
an obtaining process of obtaining intermediate images in an intermediate layer of a neural network that recognizes a quantity of objects contained in an image; and
a generating process of generating a combination image by combining together the obtained intermediate images to which a weight has been applied based on a change in recognition of the quantity of objects corresponding to a process applied to each of the intermediate images.