Apparatus for separating feature points for each object, method for separating feature points for each object and computer program

1. An object-specific keypoint separation apparatus comprising:

a processor; and

a storage medium having computer program instructions stored thereon, when executed by the processor, perform to:

receive a captured image capturing an object as an input and use a pre-trained model that has been trained in order to output a plurality of first maps and a plurality of second maps generated from the input captured image to output the plurality of first maps and the plurality of second maps, the plurality of first maps storing a vector describing a connection relationship of a keypoint of the object only around the keypoint, and the plurality of second maps representing a heat map configured to have a peak at coordinates at which the keypoint of the object appears;

correct the plurality of second maps using the plurality of first maps and the plurality of second maps output from the inference execution unit;

upsample the plurality of first maps; and

separate keypoints for each object based on the plurality of upsampled first maps and the plurality of corrected second maps.

2. The object-specific keypoint separation apparatus according to claim 1 ,

wherein the plurality of first maps and the plurality of second maps output have a low resolution, and wherein the computer program instructions further perform to upsamples the plurality of second maps to cause the plurality of second maps to have an equal resolution.

3. The object-specific keypoint separation apparatus according to claim 2 , wherein the computer program instructions further uses coordinates of a keypoint indicated by a vector on the first map to generate a vector density map representing a degree of vector density, and normalizes the vector density map by dividing a value on the vector density map by a maximum value on the generated vector density map.

4. The object-specific keypoint separation apparatus according to claim 3 , wherein the computer program instructions further perform to multiplies the normalized vector density map and each of the plurality of upsampled second maps by a predetermined percentage value and adds up multiplication results to generate the plurality of corrected second maps.

5. An object-specific keypoint separation method comprising:

receiving a captured image capturing an object as an input and using a pre-trained model that has been trained in order to output a plurality of first maps and a plurality of second maps generated from the input captured image to output the plurality of first maps and the plurality of second maps, the plurality of first maps storing a vector describing a connection relationship of a keypoint of the object only around the keypoint, and the plurality of second maps representing a heat map configured to have a peak at coordinates at which the keypoint of the object appears;

correcting the plurality of second maps using the plurality of output first maps and the plurality of output second maps;

upsampling the plurality of output first maps; and

separating the keypoints for each object based on the plurality of upsampled first maps and the plurality of corrected second maps.

6. A non-transitory computer-readable medium having computer-executable instructions that, upon execution of the instructions by a processor of a computer, cause the computer to function as the object-specific keypoint separation apparatus according to claim 1 .