Geometry coordinate scaling for AI-based dynamic point cloud coding

1 . A method of encoding point cloud data, the method comprising:

determining a number of times to downscale a representation of a point cloud geometry;

downscaling the representation of the point cloud geometry the number of times to form a downscaled representation of the point cloud geometry;

encoding the downscaled representation of the point cloud geometry;

forming encoding data representing the number of times the representation of the point cloud geometry was downscaled; and

outputting the encoded downscaled representation of the point cloud geometry and the encoded data representing the number of times the representation of the point cloud geometry was downscaled.

2 . The method of claim 1 , wherein determining the number of times to downscale the representation comprises determining the number of times to downscale the representation according to at least one of a use case, bit requirements, or a type of point cloud for the point cloud geometry.

3 . The method of claim 1 , wherein encoding the downscaled representation of the point cloud geometry comprises losslessly encoding the downscaled representation of the point cloud geometry.

4 . The method of claim 1 , further comprising encoding features corresponding to the downscaled representation of the point cloud geometry using lossy encoding.

5 . The method of claim 1 , wherein encoding the downscaled representation of the point cloud geometry comprises encoding the downscaled representation of the point cloud geometry using one of intra-frame (I-frame) encoding, inter-frame (P-frame) encoding, or bi-directional inter-frame (B-frame) encoding.

6 . The method of claim 1 , wherein the downscaled representation of the point cloud geometry comprises a downscaled representation of a point cloud geometry for a current frame, the method further comprising encoding features corresponding to the downscaled representation of the point cloud geometry, comprising:

extracting multiscale features from a previously coded frame of point cloud data;

forming, by a predictor network, predicted features of the current frame from the extracted multiscale features and the downscaled representation of the point cloud geometry for the current frame; and

coding residual data representing differences between features of the current frame and the predicted features of the current frame.

7 . The method of claim 1 , wherein downscaling the representation of the point cloud geometry and encoding the downscaled representation comprises executing a neural network trained to downscale and encode the point cloud geometry in a number of stages equal to the number of times to downscale the representation of the point cloud geometry, each stage including a downscaling step and an encoding portion.

8 . The method of claim 1 , wherein downscaling the representation of the point cloud geometry and encoding the downscaled representation comprises executing a neural network trained to downscale the point cloud geometry in a number of stages equal to the number of times to downscale the representation of the point cloud geometry, each stage including a downscaling step, and to then encode the downscaled representation of the point cloud geometry.

9 . A device for encoding point cloud data, the device comprising:

a memory configured to store point cloud data; and

one or more processors implemented in circuitry and configured to:

determine a number of times to downscale a representation of a point cloud geometry;

downscale the representation of the point cloud geometry the number of times to form a downscaled representation of the point cloud geometry;

encode the downscaled representation of the point cloud geometry;

form encoding data representing the number of times the representation of the point cloud geometry was downscaled; and

output the encoded downscaled representation of the point cloud geometry and the encoded data representing the number of times the representation of the point cloud geometry was downscaled.

10 . The device of claim 9 , wherein to determine the number of times to downscale the representation, the one or more processors are configured to determine the number of times to downscale the representation according to at least one of a use case, bit requirements, or a type of point cloud for the point cloud geometry.

11 . The device of claim 9 , wherein to encode the downscaled representation of the point cloud geometry, the one or more processors are configured to losslessly encode the downscaled representation of the point cloud geometry.

12 . The device of claim 9 , wherein the one or more processors are further configured to encode features corresponding to the downscaled representation of the point cloud geometry using lossy encoding.

13 . The device of claim 9 , wherein to encode the downscaled representation of the point cloud geometry, the one or more processors are configured to encode the downscaled representation of the point cloud geometry using one of intra-frame (I-frame) encoding, inter-frame (P-frame) encoding, or bi-directional inter-frame (B-frame) encoding.

14 . The device of claim 9 , wherein the downscaled representation of the point cloud geometry comprises a downscaled representation of a point cloud geometry for a current frame, and wherein the one or more processors are further configured to encode features corresponding to the point cloud geometry, wherein to encode the features, the one or more processors are configured to:

extract multiscale features from a previously coded frame of point cloud data;

execute a predictor network to form predicted features of the current frame from the extracted multiscale features and the downscaled representation of the point cloud geometry for the current frame; and

code residual data representing differences between features of the current frame and the predicted features of the current frame.

15 . The device of claim 9 , wherein to downscale the representation of the point cloud geometry and to encode the downscaled representation, the one or more processors are configured to execute a neural network trained to downscale and encode the point cloud geometry in a number of stages equal to the number of times to downscale the representation of the point cloud geometry, each stage including a downscaling step and an encoding portion.

16 . The device of claim 9 , wherein to downscale the representation of the point cloud geometry and to encode the downscaled representation, the one or more processors are configured to execute a neural network trained to downscale the point cloud geometry in a number of stages equal to the number of times to downscale the representation of the point cloud geometry, each stage including a downscaling step, and to then encode the downscaled representation of the point cloud geometry.

17 . A device for encoding point cloud data, the device comprising:

means for determining a number of times to downscale a representation of a point cloud geometry;

means for downscaling the representation of the point cloud geometry the number of times to form a downscaled representation of the point cloud geometry;

means for encoding the downscaled representation of the point cloud geometry;

means for forming encoded data representing the number of times the representation of the point cloud geometry was downscaled; and

means for outputting the encoded downscaled representation of the point cloud geometry and the encoded data representing the number of times the representation of the point cloud geometry was downscaled.