Granted Patent B2

US 12,450,946 · App. 18/185,981 · Granted Oct 21, 2025

Inventors: Chitranjan Jangid (Hyderabad, IN); Sandeep Sethia (Bengaluru, IN); Ravichandra Ponaganti (Hyderabad, IN)

Assignee: QUALCOMM Incorporated

G06V40/197G06T7/10G06T7/50G06T2207/20132

Abstract

Claims (100)

1. An apparatus for generating an image, the apparatus comprising:

at least one memory; and

at least one processor coupled to the at least one memory and configured to:

obtain a first image including a first subject and a second subject;

obtain a second image including the first subject and the second subject;

determine a gaze of the first subject in the first image;

maintain the first subject in the second image based on the gaze of the first subject;

obtain a depth estimate of the first subject in the first image using a depth-estimation algorithm;

obtain a depth estimate of the second subject in the first image using the depth-estimation algorithm; and

remove the second subject from the second image based on the depth estimate of the second subject being outside a depth threshold from the depth estimate of the first subject.

2. The apparatus of claim 1 , wherein the at least one processor is further configured to:

identify the first subject as a subject-of-interest of the second image based on the gaze of the first subject in the first image being toward an image sensor which captured the first image; and

remove the second subject from the second image based on identifying the first subject as a subject-of-interest of the second image.

3. The apparatus of claim 1 , wherein the at least one processor is further configured to:

identify the first subject as a subject-of-interest of the second image based on the gaze of the first subject in the first image being toward an image sensor which captured the first image;

wherein removing the second subject from the second image is further based on identifying the first subject as a subject-of-interest of the second image.

4. The apparatus of claim 1 , wherein the at least one processor is further configured to:

determine a gaze of the second subject in the first image; and

remove the second subject from the second image based on the gaze of the second subject in the first image being away from an image sensor which captured the first image.

5. The apparatus of claim 1 , wherein:

the first image includes a third subject;

the second image includes the third subject; and

the at least one processor is further configured to:

obtain a depth estimate of the first subject in the first image using a depth-estimation algorithm;

obtain a depth estimate of the third subject in the first image using the depth-estimation algorithm; and

maintain the third subject in the second image based on the depth estimate of the third subject being within a depth threshold of the depth estimate of the first subject.

6. The apparatus of claim 5 , wherein a gaze of the third subject in the first image is away from an image sensor which captured the first image.

7. The apparatus of claim 1 , wherein:

the second image is obtained from an image sensor of a device responsive to a user input instructing the device to capture the second image; and

the first image is obtained from the image sensor prior to receiving the user input.

8. The apparatus of claim 1 , wherein the at least one processor is configured to, in determining the gaze of the first subject in the first image, determine the gaze of the first subject in the first image using a trained gaze-estimation model.

9. The apparatus of claim 1 , wherein the at least one processor is further configured to replace the second subject in the second image with background pixels.

10. The apparatus of claim 9 , wherein the at least one processor is configured to, in replacing the second subject in the second image:

detect pixels representing the second subject in the second image using a subject-detection model; and

replace the pixels representing the second subject with the background pixels.

11. The apparatus of claim 9 , wherein the at least one processor is further configured to:

identify a background of the second image; and

replicate pixels from the background of the second image as the background pixels.

12. The apparatus of claim 9 , wherein the at least one processor is further configured to:

identify a background of the second image;

obtain a third image of a background of the second image; and

obtain the background pixels from the third image.

13. The apparatus of claim 12 , wherein the third image is obtained from a remote server.

14. The apparatus of claim 1 , wherein the at least one processor is further configured to:

obtain a third image including a third subject;

obtain a fourth image including the third subject;

identify a gaze of the third subject in the third image; and

add the third subject to the second image.

15. The apparatus of claim 14 , wherein:

the first image is obtained from a first image sensor of a device;

the second image is obtained from the first image sensor;

the third image is obtained from a second image sensor of the device; and

the fourth image is obtained from the second image sensor.

16. The apparatus of claim 15 , wherein the first image sensor is at a first side of the device and the second image sensor is at a second side of the device.

17. The apparatus of claim 14 , wherein:

the first image is obtained from a first device;

the second image is obtained from the first device;

the third image is obtained from a second device; and

the fourth image is obtained from the second device.

18. A method for generating an image, the method comprising:

obtaining a first image including a first subject and a second subject;

obtaining a second image including the first subject and the second subject;

determining a gaze of the first subject in the first image;

maintaining the first subject in the second image based on the gaze of the first subject;

obtaining a depth estimate of the first subject in the first image using a depth-estimation algorithm;

obtaining a depth estimate of the second subject in the first image using the depth-estimation algorithm; and

removing the second subject from the second image based on the depth estimate of the second subject being outside a depth threshold from the depth estimate of the first subject.

19. The method of claim 18 , further comprising:

identifying the first subject as a subject-of-interest of the second image based on the gaze of the first subject in the first image being toward an image sensor which captured the first image;

wherein removing the second subject from the second image is based on identifying the first subject as a subject-of-interest of the second image.

20. The method of claim 18 , further comprising:

identifying the first subject as a subject-of-interest of the second image based on the gaze of the first subject in the first image being toward an image sensor which captured the first image;

wherein removing the second subject from the second image is further based on identifying the first subject as a subject-of-interest of the second image.

21. The method of claim 18 , further comprising:

determining a gaze of the second subject in the first image;

wherein removing the second subject from the second image is based on the gaze of the second subject in the first image being away from an image sensor which captured the first image.

22. The method of claim 18 , wherein:

the first image includes a third subject;

the second image includes the third subject; and

the method further comprises:

obtaining a depth estimate of the first subject in the first image using a depth-estimation algorithm;

obtaining a depth estimate of the third subject in the first image using the depth-estimation algorithm; and

maintaining the third subject in the second image based on the depth estimate of the third subject being within a depth threshold of the depth estimate of the first subject.

23. The method of claim 22 , wherein a gaze of the third subject in the first image is away from an image sensor which captured the first image.

24. The method of claim 18 , wherein:

the second image is obtained from an image sensor of a device responsive to a user input instructing the device to capture the second image; and

the first image is obtained from the image sensor prior to receiving the user input.

25. The method of claim 18 , wherein determining the gaze of the first subject in the first image comprises determining the gaze of the first subject in the first image using a trained gaze-estimation model.

26. The method of claim 18 , further comprising replacing the second subject in the second image with background pixels.

27. The method of claim 26 , wherein replacing the second subject in the second image comprises:

detecting pixels representing the second subject in the second image using a subject-detection model; and

replacing the pixels representing the second subject with the background pixels.

28. The method of claim 26 , further comprising:

identifying a background of the second image; and

replicating pixels from the background of the second image as the background pixels.

29. The method of claim 26 , further comprising:

identifying a background of the second image;

obtaining a third image of a background of the second image; and

obtaining the background pixels from the third image.

30. The method of claim 29 , wherein the third image is obtained from a remote server.

Assignments (1)

Continuity (1)

References Cited (27)