Information processing device, information processing method, and program

Abstract

An information processing device (100) comprises a subject image acquisition unit (101) and a position estimation unit (110). The subject image acquisition unit (101) acquires an estimation subject image which includes an image of a subject site of an estimation subject. The position estimation unit (110), on the basis of a learning result obtained by performing learning using a reference image including an image of the subject site, and a subject image including an image of the subject site, extracts a subject site feature pertaining to a part corresponding to the subject site from a feature of the estimation subject image, and estimates the position of the subject site on the basis of the subject site feature.

Claims

1. A target image acquisition means for acquiring an estimated target image that includes an image of the target part of the estimated target, An extraction means that generates similarity information indicating the similarity between a reference image containing an image of the target area and the feature quantities of the estimated target image, and extracts target area feature quantities relating to the portion corresponding to the target area from the feature quantities of the estimated target image based on the similarity information, The system includes an estimation means that estimates the location of the target part based on the location of the target part by inputting the target part features into a trained position estimation model that has undergone machine learning to estimate the location of the target part in a corresponding region corresponding to the target part features, and then performs an estimation means to estimate the location of the target part based on the location of the location of the target part. Information processing device.

2. The extraction means is A generation means for generating similarity information based on the feature quantities of the image to be estimated and the reference feature quantities, Includes a feature extraction means for extracting the target part feature quantity from the estimated target image feature quantity based on the similarity information. The information processing apparatus according to claim 1.

3. The extraction means is The feature acquisition means further includes a feature acquisition means that takes the aforementioned target image as input and uses a trained feature extraction model, which has been trained to extract features from the target image, to acquire the features of the target image. The information processing apparatus according to claim 2.

4. The estimation means is, A region identification means for identifying a region position that indicates the location of the corresponding region, A first estimation means that estimates the position within the region using the position estimation model with the aforementioned target part features as input, Includes a second estimation means for estimating the position of the target part based on the aforementioned region position and the position within the region. The information processing apparatus according to any one of claims 1 to 3.

5. The estimation means is, The system further includes a conversion means for converting the estimated position of the target area to its position in the estimated target image. The information processing apparatus according to claim 4.

6. A position loss acquisition means that calculates a position loss, which is the loss related to the position within the region estimated using the position estimation model with the target part feature quantity among the features of the estimated target image as input, based on a position loss function, The system further comprises a modification means for modifying the position estimation model based on the position loss. The information processing apparatus according to any one of claims 1 to 3.

7. One or more computers, Obtain an estimated target image that includes an image of the target part in the estimated target, Similarity information is generated that shows the similarity between the reference feature quantities, which are the feature quantities of a reference image containing an image of the target area, and the feature quantities of the estimated target image. Based on the similarity information, the feature quantities of the target area corresponding to the target area are extracted from the feature quantities of the estimated target image. By inputting the target part features into a trained position estimation model that has undergone machine learning to estimate the intra-region position indicating the position of the target part in the corresponding region corresponding to the target part features, the intra-region position is estimated. The position of the target part is estimated based on the position within the aforementioned region. Information processing methods.

8. On one or more computers, Obtain an estimated target image that includes an image of the target part in the estimated target, Similarity information is generated that shows the similarity between the reference feature quantities, which are the feature quantities of a reference image containing an image of the target area, and the feature quantities of the estimated target image. Based on the similarity information, the feature quantities of the target area corresponding to the target area are extracted from the feature quantities of the estimated target image. By inputting the target part features into a trained position estimation model that has undergone machine learning to estimate the intra-region position indicating the position of the target part in the corresponding region corresponding to the target part features, the intra-region position is estimated. A program for causing the system to estimate the position of the target part based on its position within the aforementioned region.

Images