System and method for reconstructing unoccupied 3D space
The present teaching relates to method, system, medium, and implementations for understanding a three dimensional (3D) scene. Image data acquired by a camera at different time instances with respect to the 3D scene are received wherein the 3D scene includes a user or one or more objects. The face of the user is detected and tracked at different time instances. With respect to some of the time instances, a 2D user profile representing a region in the image data occupied by the user is generated based on a corresponding face detected and a corresponding 3D space in the 3D scene is estimated based on calibration parameters associated with the camera. Such estimated 3D space occupied by the user in the 3D scene is used to dynamically update a 3D space occupancy record of the 3D scene.
1. A method implemented on at least one machine including at least one processor, memory, and communication platform capable of connecting to a network for understanding a three dimensional (3D) scene, the method comprising:
receiving image data acquired, by a camera at different time instances, with respect to the 3D scene which includes at least one of a user and one or more objects present therein;
detecting a face of the user at each of the different time instances; and
with respect to each of at least some of the different time instances,
generating a 2D user profile of the user based on the face detected at the time instance, wherein the 2D user profile represents a region in the image data occupied by the user at the time instance,
obtaining a 3D position of the face of the user in the 3D scene corresponding to the region in the image data,
determining, based on a face based human model, a 3D prism in the 3D scene at the 3D position,
estimating a 3D volumetric space occupied by the user in the 3D scene based on the 3D prism,
determining spatial relationships between the one or more objects and the 3D volumetric space,
estimating, based on the 3D volumetric space and the spatial relationships, a 3D space occupancy of the 3D scene, and
dynamically updating a 3D space occupancy record of the 3D scene to reflect the estimated 3D space occupancy of the 3D scene.
2. The method of claim 1 , wherein the user is engaged in a human machine dialogue.
3. The method of claim 2 , wherein the 3D scene is a dialogue scene where the user is engaged in the human machine dialogue.
4. The method of claim 1 , further comprising:
detecting, from the image data, the one or more objects; and
estimating at least one feature associated with each of the one or more objects.
5. The method of claim 4 , further comprising determining a 3D pose of each of the one or more objects.
6. The method of claim 5 , further comprising updating the 3D space occupancy record based on the 3D pose of each of the detected one or more objects.
7. The method of claim 1 , further comprising determining a trajectory of the user in the 3D scene based on the at least some 3D spaces estimated based on detected user face.
8. Machine readable and non-transitory medium having information recorded thereon for understanding a three dimensional (3D) scene, wherein the information, when read by the machine, causes the machine to perform:
receiving image data acquired, by a camera at different time instances, with respect to the 3D scene which includes at least one of a user and one or more objects present therein;
detecting a face of the user at each of the different time instances; and
with respect to each of at least some of the different time instances,
generating a 2D user profile of the user based on the face detected at the time instance, wherein the 2D user profile represents a region in the image data occupied by the user at the time instance,
obtaining a 3D position of the face of the user in the 3D scene corresponding to the region in the image data,
determining, based on a face based human model, a 3D prism in the 3D scene at the 3D position,
estimating a 3D volumetric space occupied by the user in the 3D scene based on the 3D prism,
determining spatial relationships between the one or more objects and the 3D volumetric space,
estimating, based on the 3D volumetric space and the spatial relationships, a 3D space occupancy of the 3D scene, and
dynamically updating a 3D space occupancy record of the 3D scene to reflect the estimated 3D space occupancy of the 3D scene.
9. The medium of claim 8 , wherein the user is engaged in a human machine dialogue.
10. The medium of claim 9 , wherein the 3D scene is a dialogue scene where the user is engaged in the human machine dialogue.
11. The medium of claim 8 , wherein the information, when read by the machine, further causes the machine to perform:
detecting, from the image data, the one or more objects; and
estimating at least one feature associated with each of the one or more objects.
12. The medium of claim 11 , wherein the information, when read by the machine, further causes the machine to perform determining a 3D pose of each of the one or more objects.
13. The medium of claim 12 , wherein the information, when read by the machine, further causes the machine to perform updating the 3D space occupancy record based on the 3D pose of each of the detected one or more objects.
14. The medium of claim 8 , wherein the information, when read by the machine, further causes the machine to perform determining a trajectory of the user in the 3D scene based on the at least some 3D spaces estimated based on detected user face.
15. A system for understanding a three dimensional (3D) scene, comprising:
a face detection unit configured for
receiving image data acquired, by a camera at different time instances, with respect to the 3D scene which includes at least one of a user and one or more objects present therein, and
detecting a face of the user at each of the different time instances; and
a faced based human tracking unit configured for, with respect to each of at least some of the different time instances, generating a 2D user profile of the user based on the face detected at the time instance, wherein the 2D user profile represents a region in the image data occupied by the user at the time instance;
a human 3D occupancy estimator configured for, with respect to the at least some of the time instances,
obtaining a 3D position of the face of the user in the 3D scene corresponding to the region in the image data,
determining, based on a face based human model, a 3D prism in the 3D scene at the 3D position,
estimating a 3D volumetric space occupied by the user in the 3D scene based on the 3D prism, and
a spatial relationship identifier configured for determining spatial relationships between the one or more objects and the 3D volumetric space; and
a dynamic occupancy updater configured for dynamically updating a 3D space occupancy record of the 3D scene to reflect a 3D space occupancy of the 3D scene that is estimated based on the 3D volumetric space and the spatial relationships.
16. The system of claim 15 , wherein the user is engaged in a human machine dialogue.
17. The system of claim 16 , wherein the 3D scene is a dialogue scene where the user is engaged in the human machine dialogue.
18. The system of claim 15 , further comprising:
a visual object detection unit configured for detecting, from the image data, the one or more objects; and
an object property determiner configured for estimating at least one feature associated with each of the one or more objects.
19. The system of claim 18 , further comprising an object pose estimator configured for determining a 3D pose of each of the one or more objects.
20. The system of claim 19 , wherein the dynamic occupancy updater is further configured for updating the 3D space occupancy record based on the 3D pose of each of the detected one or more objects.
21. The system of claim 15 , further comprising a face based human tracking unit configured for determining a trajectory of the user in the 3D scene based on the at least some 3D spaces estimated based on detected user face.