Image segmentation model training method and electronic device
The disclosure provides an image segmentation model training method and an electronic device. The method includes the following steps. A plurality of labeled images of a training image are obtained, wherein each of the labeled images includes a labeled target area and a background area. A union area of the labeled target areas of the plurality of labeled images is obtained, and a complement area of the union area is obtained. A predicted segmented image of the training image is generated using an image segmentation model. A first loss value is generated according to the predicted segmented image and the complement area. A second loss value is generated according to the predicted segmented image and the labeled target area of at least one of the plurality of labeled images. The image segmentation model is trained according to the first loss value and the second loss value.
1 . An image segmentation model training method performed by an electronic device and comprising:
obtaining a plurality of labeled images of a training image, wherein each of the plurality of labeled images comprises a labeled target area and a background area;
obtaining a union area of the labeled target areas of the plurality of labeled images, and obtaining a complement area of the union area;
generating a predicted segmented image of the training image utilizing an image segmentation model;
generating a first loss value according to the predicted segmented image and the complement area;
generating a second loss value according to the predicted segmented image and the labeled target area of at least one of the plurality of labeled images; and
training the image segmentation model according to the first loss value and the second loss value.
2 . The image segmentation model training method of claim 1 , wherein the predicted segmented image comprises a predicted target area and a predicted background area, and the step of generating the first loss value according to the predicted segmented image and the complement area comprises:
generating the first loss value according to a degree of overlap between the predicted background area of the predicted segmented image and the complement area.
3 . The image segmentation model training method of claim 2 , wherein the step of generating the first loss value according to the degree of overlap of the predicted background area of the predicted segmented image and the complement area comprises:
generating a first coefficient according to the degree of overlap between the predicted background area of the predicted segmented image and the complement area; and
generating the first loss value according to a difference between the first coefficient and a default value.
4 . The image segmentation model training method of claim 3 , wherein the first coefficient is a Dice coefficient.
5 . The image segmentation model training method of claim 1 , wherein the step of generating the second loss value according to the predicted segmented image and the labeled target area of at least one of the plurality of labeled images comprises:
obtaining an intersection area of the labeled target areas of the plurality of labeled images; and
generating the second loss value according to the predicted segmented image and the intersection area.
6 . The image segmentation model training method of claim 5 , wherein the predicted segmented image comprises a predicted target area and a predicted background area, and the step of generating the second loss value according to the predicted segmented image and the intersection area comprises:
generating the second loss value according to a degree of overlap between the predicted target area of the predicted segmented image and the intersection area.
7 . The image segmentation model training method of claim 6 , wherein the step of generating the second loss value according to the degree of overlap of the predicted target area of the predicted segmented image and the intersection area comprises:
generating a second coefficient according to a degree of overlap between the predicted background area of the predicted segmented image and the intersection area; and
generating the second loss value according to a difference between the second coefficient and a default value.
8 . The image segmentation model training method of claim 7 , wherein the second coefficient is a Dice coefficient.
9 . The image segmentation model training method of claim 5 , wherein the step of training the image segmentation model according to the first loss value and the second loss value comprises:
performing a summing process on the first loss value and the second loss value to obtain a target loss value; and
training the image segmentation model according to the target loss value.
10 . The image segmentation model training method of claim 1 , wherein the training image is a medical image, and the image segmentation model is a deep learning model.
11 . An electronic device comprising:
a storage device storing a program code; and
a processor coupled to the storage device and accessing the program code to execute:
obtaining a plurality of labeled images of a training image, wherein each of the plurality of labeled images comprises a labeled target area and a background area;
obtaining a union area of the labeled target areas of the plurality of labeled images, and obtaining a complement area of the union area;
generating a predicted segmented image of the training image utilizing an image segmentation model;
generating a first loss value according to the predicted segmented image and the complement area;
generating a second loss value according to the predicted segmented image and the labeled target area of at least one of the plurality of labeled images; and
training the image segmentation model according to the first loss value.
12 . The electronic device of claim 11 , wherein the predicted segmented image comprises a predicted target area and a predicted background area, and the processor executes:
generating the first loss value according to a degree of overlap between the predicted background area of the predicted segmented image and the complement area.
13 . The electronic device of claim 12 , wherein the processor executes:
generating a first coefficient according to the degree of overlap between the predicted background area of the predicted segmented image and the complement area; and
generating the first loss value according to a difference between the first coefficient and a default value.
14 . The electronic device of claim 13 , wherein the first coefficient is a Dice coefficient.
15 . The electronic device of claim 11 , wherein the processor executes:
obtaining an intersection area of the labeled target areas of the plurality of labeled images;
generating the second loss value according to the predicted segmented image and the intersection area.
16 . The electronic device of claim 15 , wherein the predicted segmented image comprises a predicted target area and a predicted background area, and the processor executes:
generating the second loss value according to a degree of overlap between the predicted target area of the predicted segmented image and the intersection area.
17 . The electronic device of claim 16 , wherein the processor executes:
generating a second coefficient according to a degree of overlap between the predicted background area of the predicted segmented image and the intersection area; and
generating the second loss value according to a difference between the second coefficient and a default value.
18 . The electronic device of claim 17 , wherein the second coefficient is a Dice coefficient.
19 . The electronic device of claim 15 , wherein the processor executes:
performing a summing process on the first loss value and the second loss value to obtain a target loss value; and
training the image segmentation model according to the target loss value.
20 . The electronic device of claim 11 , wherein the training image is a medical image, and the image segmentation model is a deep learning model.