DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
Applicant’s response, filed 2 January 2026, to the last office action has been entered and made of record.
In response to the amendments to the specification and claims, they are acknowledged, supported by the original disclosure, and no new matter is added.
In response to the amendments to the specification, the amended language has overcome the objections to the specification of the previous Office action, and the respective objections have been withdrawn.
In response to the amendments to the claims, specifically addressing the treatment of claims 1-10 under 35 U.S.C. § 112(f) / (pre-AIA ) sixth paragraph, of the previous Office action, the amended language has obviated necessity to invoke 35 U.S.C. § 112(f) / (pre-AIA ) sixth paragraph, and the respective claims are treated with their broadest reasonable interpretation, in light of the specification.
In response to the amendments to the claims, specifically addressing the rejections of claims 12 and 14 under 35 U.S.C. § 101, of the previous Office action, the amended language has overcome the respective rejections, and the rejections have been withdrawn.
Amendments to the independent claims 1, 6, 11-14 have necessitated updated ground of rejection over the applied prior art. Please see below for the updated interpretations and rejections.
In response to the addition of new claims 15 and 16, they are acknowledged and made of record.
Response to Arguments
Applicant's arguments filed 2 January 2026 have been fully considered but they are not persuasive.
Examiner notes the claims are treated with their broadest reasonable interpretations consistent with the specification. See MPEP 2111. Although the claims are interpreted in light of the specification, limitations from the specification are not read into the claims. See In re Van Geuns, 988 F.2d 1181, 26 USPQ2d 1057 (Fed. Cir. 1993). Furthermore, the test for obviousness is what the combined teachings of the references would have suggested to those of ordinary skill in the art. See In re Keller, 642 F.2d 413, 208 USPQ871 (CCPA 1981).
In response to Applicant’s arguments beginning on p. 15 of Applicant’s reply, that the combined cited prior art teachings, notably Cao, fails to fairly teach or suggest performing motion estimation between a reference frame of fractional precision and a frame of integer precision to generate the motion information, the Examiner respectfully disagrees.
Yea is relied upon to teach a method and device for interframe point cloud attribute coding in performing compression and decompression of point cloud data, where received point clouds are encoded with an octree encoding algorithm, generating a LoD of each of the points corresponding to received reordered positions, and predicting residuals respectively of received attributes of points in the input point cloud by applying a prediction algorithm, such as interpolation, weighted average calculation, etc., to the received attributes in an order based on received LoD of each of the points corresponding to reordered positions from the octree encoder (see Yea Fig. 1 and [0007]-[0011]; Fig. 4 and [0037]-[0047]).
Chou’100 is relied upon to teach a known technique of compressing the geometry data of a point cloud using a octree scanning of points in 3D space, where the point cloud data is hierarchically organized in levels in 3D space, with a WxWxW unit cube for a voxel being partitioned into eight smaller sub cubes W/2xW/2xW/2, which can be further partitioned into smaller sub cubes (see Chou’100 Fig. 3, [0046], [0049], [0054]-[0055], Fig. 9, and [0107]-[0109]).
The combined teachings of Yea and Chou’100 suggests to one of ordinary skill in the art that the geometry data of a point cloud data are encoded with an octree frame with sub cube partitions, providing for fractional precision in encoding positions of the point cloud data.
Yea and Chou’100 further suggests that geometry based or joint geometry/attribute based global/local motion estimation may be performed for motion compensation when determining prediction values (see Yea [0060]-[0063] and Chou’100 [0060]-[0061]).
Cao is noted to be further relied upon to teach a known technique of performing global and local motion estimation between a prediction frame and current frame to perform motion compensation and encode motion vector information and point information, where local node motion estimation is applied to estimate a local motion at a finer scale and a node level in an octree, where local node motion vector of a node of the octree is a motion vector that is only applied on points within the node in prediction (reference) frame, and the local node motion vector is determined in a recursive manner using a cost function to choose the best suitable motion vector, and a respective motion vector can be assigned to sub-nodes if the current node is divided into children nodes (see Cao Fig. 4, Fig. 6, [0083]-[0084], and [0088]-[0089]).
The combined teachings of Yea and Chou’100 with the teachings of Cao suggests to one of ordinary skill in the art that local node motion estimation is performed to estimate a local motion at a finer scale and a node level in an octree by determining local node motion vectors of a node of the octree, which includes sub cube partitions with fractional precision, that is applied on points within the node in reference frame.
Thus, the combined cited prior art teachings provides for the broadest reasonable interpretation for the claimed subject matter of performing motion estimation between a reference frame of fractional precision and a frame of integer precision to generate the motion information.
In response to Applicant’s arguments beginning on p. 18 of Applicant’s reply, that the combined cited prior art teachings, notably Yea, fails to fairly teach or suggest performing an interpolation process on a reconstructed point cloud of a decoded frame with respect to attribute information in point cloud information to generate a reference frame of fractional precision, the Examiner respectfully disagrees.
Yea is noted to be relied upon to teach a G-PCC decompressor, where compressed bitstream is received and decoded to obtain occupancy symbols and quantized positions using an octree decoding algorithm, obtains levels of detail (LoD) of each of the points corresponding to the quantized positions, and obtain reconstructed prediction residuals, and reconstructs attributes respective of the reconstructed prediction residuals using a prediction algorithm, such as interpolation or weighted average calculations (see Yea Fig. 5 and [0051]-[0057]).
Chou is further relied upon to teach receiving and decompressing compressed geometry data, where the point cloud data is hierarchically organized in levels in 3D space, with a WxWxW unit cube for a voxel being partitioned into eight smaller sub cubes W/2xW/2xW/2, which can be further partitioned into smaller sub cubes (see Chou’100 Fig. 4 and [0076], Fig. 9, and [0107]-[0109]).
The combined teachings of Yea and Chou’100 suggests to one of ordinary skill in the art that the geometry and attribute data of a point cloud data are reconstructed from decoding a compressed bitstream, where the point cloud data is organized with an octree frame with sub cube partitions, providing for fractional precision, and the attributes are reconstructed using an interpolation based prediction algorithm.
Thus, the combined cited prior art teachings provides for the broadest reasonable interpretation for the claimed subject matter of performing an interpolation process on a reconstructed point cloud of a decoded frame with respect to attribute information in point cloud information to generate a reference frame of fractional precision.
In response to Applicant’s arguments beginning on p. 19 of Applicant’s reply, that the new independent claims are allowable over the cited prior art references, the Examiner respectfully disagrees.
Examiner notes that the cited Yea and Chou’120 references provides further teachings relevant to the further recited features of determining neighboring points among integer-precision voxels when a distance between coordinates of two voxels is below a threshold (p), and performs the interpolation process only between such neighboring voxels to generate fractional-precision geometry and attribute values.
Notably, Yea further teaches that the prediction algorithm may include interpolation, weighted average calculation, a nearest neighbor algorithm, and RDO (see Yea Fig. 4 and [0046]-[0047]), and that a nearest neighbor search is performed for each point of the point cloud, and can be conducted both at the encoder and the decoder (see Yea [0064]-[0073]).
Chou’120 is further noted to teach that a threshold is used upon the distance between voxels i and j to control the size of the neighborhood of a voxel for filtering geometry elements to smooth the surface (see Chou’120 [0160]).
The combination of the further teachings of Yea and Chou’120 suggests to one of ordinary skill in the art that a neighborhood size can be controlled by comparing distances between received attribute points when performing the neighborhood search for applying the prediction algorithm.
Thus, the combined cited prior art teachings provides for the broadest reasonable interpretation for the additional claimed subject matter of new claims 15-16. Please see below for relevant updated prior art rejections.
Applicant’s remaining arguments for the remaining claims refer to the arguments presented to the claims discussed above, the Examiner similarly refers to the corresponding discussion above.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
Claims 1-16 are rejected under 35 U.S.C. 103 as being unpatentable over Yea et al. (US 2020/0304829), herein Yea, in view of Chou et al. (US 2017/0347100), herein Chou’100, Cao et al. (US 2022/0108487, effectively filed 6 October 2021), herein Cao, and Chou et al. (US 2017/0347120), herein Chou’120.
Regarding claim 1, Yea discloses a point cloud coding device comprising:
interpolation circuitry (see Yea [0095], [0109]-[0110], [0119]-[0122], where the disclosed teachings are implemented by processors executing programs stored in non-transitory computer readable mediums) configured to perform an interpolation process on a reconstructed point cloud of a coded frame with respect to attribute information indicating an attribute of a point cloud (see Yea Fig. 1 and [0007]-[0011], where in graph based point cloud compression (G-PCC), levels of detail (LoD) of each 3D point are generated based on a distance of each 3D point and attributes values of 3D points in each LoD is encoded by applying prediction in an LoD based order, where if a variability of a neighborhood of a 3D point is lower than a threshold, the distance based weighted average prediction is conducted by predicting attribute values using a linear interpolation process based on distances of nearest neighbors of current point; see Yea Fig. 4 and [0037]-[0047], where the G-PCC compressor receives positions of points in an input point cloud, an octree encoder encodes filtered and quantized positions into occupancy symbols of an octree representing the input point cloud using an octree encoding algorithm and produces reordered positions, a prediction module obtains prediction residuals respectively of received attributes of points in the input point cloud by applying a prediction algorithm to the received attributes in an order based on received LoD of each of the points corresponding to reordered positions from the octree encoder, where the prediction algorithm may include interpolation, weighted average calculation, a nearest neighbor algorithm and RDO).
While Yea teaches that received point clouds are encoded with an octree encoding algorithm and generating a LoD of each of the points corresponding to received reordered positions (see Yea Fig. 4 and [0037]-[0047]);
Yea does not explicitly disclose generate a reference frame of fractional precision.
Chou’100 teaches in a pertinent and related compression and decompression of point cloud data (see Chou’100 Abstract), where received point cloud data includes geometry data for points as well as attributes of occupied points, and an octree coder compresses the geometry data using a octree scanning of points in 3D space, where the point cloud data is hierarchically organized in levels in 3D space, with a WxWxW unit cube for a voxel being partitioned into eight smaller sub cubes W/2xW/2xW/2, which can be further partitioned into smaller sub cubes (see Chou’100 Fig. 3, [0046], [0049], [0054]-[0055], Fig. 9, and [0107]-[0109]).
At the time of filing, one of ordinary skill in the art would have found it obvious to apply the teachings of Chou’100 to the teachings of Yea such that the geometry data of a point cloud data are encoded with an octree frame with sub cube partitions, providing for fractional precision in encoding positions of the point cloud data.
This modification is rationalized as an application of a known technique to a known device ready for improvement to yield predictable results.
In this instance, Yea disclose a base method and device for interframe point cloud attribute coding in performing compression and decompression of point cloud data, where received point clouds are encoded with an octree encoding algorithm, generating a LoD of each of the points corresponding to received reordered positions, and predicting residuals respectively of received attributes of points in the input point cloud by applying a prediction algorithm, such as interpolation, weighted average calculation, etc., to the received attributes in an order based on received LoD of each of the points corresponding to reordered positions from the octree encoder.
Chou’100 teaches a known technique of compressing the geometry data of a point cloud using a octree scanning of points in 3D space, where the point cloud data is hierarchically organized in levels in 3D space, with a WxWxW unit cube for a voxel being partitioned into eight smaller sub cubes W/2xW/2xW/2, which can be further partitioned into smaller sub cubes.
One of ordinary skill in the art would have recognized that by applying Chou’100’s techniques would allow for the octree encoding performed by Yea’s device to encode the geometry data of a point cloud data with an octree frame with sub cube partitions with fractional precision when encoding positions of the point cloud data, predictably leading to an improved point cloud compression with improved geometric encoding precision of point cloud data.
While Yea and Chou’100 teach that a geometry based or joint geometry/attribute based global/local motion estimation may be performed for motion compensation when determining prediction values (see Yea [0060]-[0063]), and estimating the motion of attributes of the block with respect to one or more reference frames of point cloud data (see Chou’100 [0060]-[0061);
Yea and Chou’100 do not explicitly disclose motion estimation circuitry configured to perform motion estimation between the reference frame of fractional precision and a frame of integer precision to generate motion information; and prediction circuitry configured to generate a predicted value based on the motion information.
Cao teaches in a related and pertinent device and method for encoding point clouds (see Cao Abstract), where, given a prediction frame and current frame, global and local motion estimation is performed and applied to perform motion compensation and encode motion vector information and point information where local node motion estimation is applied to estimate a local motion at a finer scale and a node level in an octree (see Cao Fig. 4 and [0084]), where local node motion vector of a node of the octree is a motion vector that is only applied on points within the node in prediction (reference) frame (see Cao [0083]), and a G-PCC encoder may determine the local node motion vector in a recursive manner using a cost function to choose the best suitable motion vector, and a respective motion vector can be assigned to sub-nodes if the current node is divided into children nodes (see Cao Fig. 6, [0088]-[0089]).
At the time of filing, one of ordinary skill in the art would have found it obvious to apply the teachings of Cao to the teachings of Yea and Chou’100 such that global and local motion estimation is performed for motion compensation when encoding the point cloud data, where local node motion estimation as taught by Cao is performed to estimate a local motion at a finer scale and a node level in an octree by determining local node motion vectors of a node of the octree, which includes sub cube partitions with fractional precision, that is only applied on points within the node in reference frame.
This modification is rationalized as an application of a known technique to a known device ready for improvement to yield predictable results.
In this instance, Yea and Chou’100 disclose a base method and device for interframe point cloud attribute coding in performing compression and decompression of point cloud data, where received point clouds are encoded with an octree encoding algorithm, where the geometry data of a point cloud data is encoded with an octree frame with sub cube partitions with fractional precision, generating a LoD of each of the points corresponding to received reordered positions, and predicting residuals respectively of received attributes of points in the input point cloud by applying a prediction algorithm, such as interpolation, weighted average calculation, etc., to the received attributes in an order based on received LoD of each of the points corresponding to reordered positions from the octree encoder, where a geometry based or joint geometry/attribute based global/local motion estimation may be performed for motion compensation when determining prediction values, and estimating the motion of attributes of the block with respect to one or more reference frames of point cloud data.
Cao teaches a known technique of performing global and local motion estimation to perform motion compensation and encode motion vector information and point information, where local node motion estimation is applied to estimate a local motion at a finer scale and a node level in an octree, where local node motion vector of a node of the octree is a motion vector that are applied on points within the node in the reference frame, and determined in a recursive manner using a cost function to choose the best suitable motion vector, and a respective motion vector can be assigned to sub-nodes when the current node is divided into children nodes.
One of ordinary skill in the art would have recognized that by applying Cao’s techniques would allow for the point cloud encoding performed by Yea and Chou’100’s device to perform global and local motion estimation for motion compensation, where local node motion estimation as taught by Cao is performed to estimate a local motion at a finer scale and a node level in an octree by determining local node motion vectors of a node of the octree, which includes sub cube partitions with fractional precision, that is applied on points within the node in reference frame, predictably leading to an improved point cloud compression with local node level motion estimation when performing motion compensation for estimating the motion of attributes of the point cloud data.
While Yea, Chou’100, and Cao teaches prediction residuals are quantized and entropy encoding is performed to obtain a compressed bitstream (see Yea [0049]; see Chou’100 [0058]; see Cao [0060] and [0064]);
Yea, Chou’100, and Cao do not explicitly disclose entropy coding circuitry configured to entropy-code a difference between a point cloud of the coded frame and the predicted value.
Chou’120 teaches in a related and pertinent method for motion compensated encoding point clouds (see Chou’120 Abstract), where motion data is estimated and provided for entropy encoding and inter-frame compression can be determined for compressing a given block, where encodes and transmits the differences between prediction values and corresponding original attributes, where the differences provide values of the residuals, and values of the prediction residual are encoded using region-adaptive hierarchical transformer, quantizer, and entropy encoder (see Chou’120 [0063]-[0068]).
At the time of filing, one of ordinary skill in the art would have found it obvious to apply the teachings of Chou’120 to the teachings of Yea, Chou’100, and Cao, such that inter-frame compression can be applied to perform motion compensated encoding of point cloud data, as taught by Chou’120, where differences between prediction values and corresponding original attributes of the point cloud data are determined, quantized and entropy encoded to obtain and output compressed bitstream.
This modification is rationalized as an application of a known technique to a known device ready for improvement to yield predictable results.
In this instance, Yea, Chou’100, and Cao disclose a base method and device for interframe point cloud attribute coding in performing compression and decompression of point cloud data, where received point clouds are encoded with an octree encoding algorithm, generating a LoD of each of the points corresponding to received reordered positions, and predicting residuals respectively of received attributes of points in the input point cloud by applying a prediction algorithm, such as interpolation, weighted average calculation, etc., to the received attributes in an order based on received LoD of each of the points corresponding to reordered positions from the octree encoder, where a geometry based or joint geometry/attribute based global/local motion estimation may be performed for motion compensation when determining prediction values, estimating the motion of attributes of the block with respect to one or more reference frames of point cloud data, and prediction residuals are quantized and entropy encoded to obtain a compressed bitstream.
Chou’120 teaches a known technique of performing motion compensated point cloud compression, where motion data is estimated and provided for entropy encoding, and inter-frame compression can be determined for compressing a given block, where encodes and transmits the differences between prediction values and corresponding original attributes, where the differences provide values of the residuals, and values of the prediction residual are encoded using region-adaptive hierarchical transformer, quantizer, and entropy encoder.
One of ordinary skill in the art would have recognized that by applying techniques would allow for the point cloud encoding performed by Yea, Chou’100, and Cao's device to apply inter-frame compression when performing motion compensated encoding of point cloud data, as taught by Chou’120, where differences between prediction values and corresponding original attributes of the point cloud data are determined, quantized and entropy encoded to obtain and output compressed bitstream, predictably leading to an improved point cloud compression with interframe motion compression of the point cloud data.
Regarding claim 2, please see the above rejection of claim 1. Yea, Chou’100, Cao, and Chou’120 discloses the point cloud coding device according to claim 1, wherein the interpolation circuitry generates an interpolated value by performing the interpolation process on attribute values based on attribute values of two closest points with respect to each fractional precision position (see Yea Fig. 1 and [0007]-[0011], where attribute values of a 3D point are predicted by using a linear interpolation process of distances of nearest neighbors of a current point, where for an exemplified 3D point P2, neighboring three points P0,P5, and P4 are suggested; see Yea Fig. 4 and [0037]-[0047], where input point clouds are octree encoded and prediction residuals are obtained respectively of received attributes of points in the input point cloud by applying a prediction algorithm to the received attributes in an order based on received LoD of each of the points corresponding to reordered positions from the octree encoder, where the prediction algorithm may include interpolation, weighted average calculation; and see Chou’100 Fig. 3, [0046], [0049], [0054]-[0055], Fig. 9, and [0107]-[0109], where point cloud data is hierarchically organized in levels in 3D space, with a WxWxW unit cube for a voxel being partitioned into eight smaller sub cubes W/2xW/2xW/2, which can be further partitioned into smaller sub cubes; where the combined teaching suggests that the 3D points at a sub cube position can be interpolated by taking a distance based weighted average of nearest neighbors of the point).
Regarding claim 3, please see the above rejection of claim 2. Yea, Chou’100, Cao, and Chou’120 disclose the point cloud coding device according to claim 2, wherein, in a case where there are two or more interpolated values with respect to each fractional precision position, the interpolation circuitry uses an average value of the two or more interpolated values as the interpolated value (see Yea Fig. 1 and [0007]-[0011], where attribute values of a 3D point are predicted by using a linear interpolation process of distances of nearest neighbors of a current point, where for an exemplified 3D point P2, neighboring three points P0,P5, and P4 are suggested; see Yea Fig. 4 and [0037]-[0047], where input point clouds are octree encoded and prediction residuals are obtained respectively of received attributes of points in the input point cloud by applying a prediction algorithm to the received attributes in an order based on received LoD of each of the points corresponding to reordered positions from the octree encoder, where the prediction algorithm may include interpolation and weighted average calculation; and see Chou’100 Fig. 3, [0046], [0049], [0054]-[0055], Fig. 9, and [0107]-[0109], where point cloud data is hierarchically organized in levels in 3D space, with a WxWxW unit cube for a voxel being partitioned into eight smaller sub cubes W/2xW/2xW/2, which can be further partitioned into smaller sub cubes; where the combined teaching suggests that the 3D points at a sub cube position can be interpolated by taking a distance based weighted average of nearest neighbors of the point, where for further partitioned sub cube positions, the neighboring points would be interpolated values and an average of nearest neighbors of interpolated values would be calculated for the further portioned sub cube point).
Regarding claim 4, please see the above rejection of claim 1. Yea, Chou’100, Cao, and Chou’120 disclose the point cloud coding device according to claim 1, wherein
the prediction circuitry uses, to generate the predicted value, an attribute value of a closest point with respect to each point in the frame of integer precision among points of the reference frame of fractional precision after motion compensation based on the motion information (see Yea Fig. 1, [0011]-[0014], and Table 1, where if variability is higher than a threshold, a rate distortion optimized predictor selection is performed where multiple predictor candidate or candidate predicted values are created based on a result of a neighbor point search in generating LoD, where attribute values of a nearest neighbor point is selected ; see Yea Fig. 4 and [0037]-[0047], where input point clouds are octree encoded and prediction residuals are obtained respectively of received attributes of points in the input point cloud by applying a prediction algorithm to the received attributes in an order based on received LoD of each of the points corresponding to reordered positions from the octree encoder, where the prediction algorithm may include interpolation, weighted average calculation; and see Chou’100 Fig. 3, [0046], [0049], [0054]-[0055], Fig. 9, and [0107]-[0109], where point cloud data is hierarchically organized in levels in 3D space, with a WxWxW unit cube for a voxel being partitioned into eight smaller sub cubes W/2xW/2xW/2, which can be further partitioned into smaller sub cubes; see Chou’100 [0060]-[0061], where the motion of attributes are estimated of the block with respect to one or more reference frames of point cloud data; where the combined teaching suggests that motion compensated attributes of integer unit cube positions may be used as a predicted attribute value for sub cube positions).
Regarding claim 5, please see the above rejection of claim 1. Yea, Chou’100, Cao, and Chou’120 disclose the point cloud coding device according to claim 1, wherein the fractional precision is 1/2 voxel precision (see Chou’100 Fig. 3, [0046], [0049], [0054]-[0055], Fig. 9, and [0107]-[0109], where the point cloud data is hierarchically organized in levels in 3D space, with a WxWxW unit cube for a voxel being partitioned into eight smaller sub cubes W/2xW/2xW/2, which can be further partitioned into smaller sub cubes).
Regarding claim 6, Yea, Chou’100, Cao, and Chou’120 disclose a point cloud decoding device comprising:
interpolation circuitry (see Yea [0095], [0109]-[0110], [0119]-[0122], where the disclosed teachings are implemented by processors executing programs stored in non-transitory computer readable mediums) configured to perform an interpolation process on a reconstructed point cloud of a decoded frame with respect to attribute information in point cloud information and generate a reference frame of fractional precision (see Yea Fig. 1 and [0007]-[0011], where in graph based point cloud compression (G-PCC), levels of detail (LoD) of each 3D point are generated based on a distance of each 3D point and attributes values of 3D points in each LoD is encoded by applying prediction in an LoD based order, where if a variability of a neighborhood of a 3D point is lower than a threshold, the distance based weighted average prediction is conducted by predicting attribute values using a linear interpolation process based on distances of nearest neighbors of current point; see Yea Fig. 5 and [0051]-[0057], where a G-PCC decompressor is disclosed, where compressed bitstream is received and decoded to obtain occupancy symbols, quantized positions using an octree decoding algorithm, obtains LoD of each of the points corresponding to the quantized positions, and obtain reconstructed prediction residuals, and reconstructs attributes respective of the reconstructed prediction residuals using a prediction algorithm, such as interpolation or weighted average calculations; see Chou’100 Fig. 4 and [0076], Fig. 9, and [0107]-[0109], where compressed geometry data is received and decompressed, where the point cloud data is hierarchically organized in levels in 3D space, with a WxWxW unit cube for a voxel being partitioned into eight smaller sub cubes W/2xW/2xW/2, which can be further partitioned into smaller sub cubes);
entropy decoding circuitry configured to decode motion information and a predicted residual from a bit stream (see Yea [0052], where compressed bitstream is received and decoded using various entropy decoding algorithms to obtain occupancy symbols and quantized prediction residuals; see Chou’120 [0077]-[0079] and [0084], where entropy decoding is performed to decode quantized transform coefficients from received encoded data, and process residual values for any of the attributes of occupied points and decode motion data);
prediction circuitry configured to generate a predicted value based on the motion information and the reference frame of fractional precision (see Yea [0057], where attributes are reconstructed respective of reconstructed prediction residuals using a prediction algorithm, such as interpolation or weighted average calculations; see Yea [0060]-[0063], where a geometry based or joint geometry/attribute based global/local motion estimation may be performed for motion compensation when determining prediction values; see Chou’100 [0081]-[0082], where decoded motion data is used to produce motion compensated prediction values; see Cao [0083], where a G-PCC coder may apply local node motion compensation only on a portion of the reference frame, where local node motion vector of a node of the octree is a motion vector that is only applied on points within the node in prediction (reference) frame); and
attribute information decoding circuitry configured to use a sum of the predicted residual and the predicted value as a decoding value of an attribute value (see Chou’120 [0086], where reconstructed residual values are combined with the prediction values to produce a reconstruction of the attributes of occupied points for the current point cloud frame).
Please see the above rejection for claim 1, as claim 6 is a corresponding decoding device to the encoding device of claim 1, and the rationale to combine the teachings of Yea, Chou’100, Cao, and Chou’120 are similar, mutatis mutandis.
Regarding claim 7, please see the above rejection of claim 6. Yea, Chou’100, Cao, and Chou’120 disclose the point cloud decoding device according to claim 6, wherein the interpolation circuitry generates an interpolated value by performing an interpolation process based on attribute values of two closest points with respect to each fractional precision position (see Yea Fig. 1 and [0007]-[0011], where attribute values of a 3D point are predicted by using a linear interpolation process of distances of nearest neighbors of a current point, where for an exemplified 3D point P2, neighboring three points P0,P5, and P4 are suggested; see Yea Fig. 5 and [0051]-[0057], where a G-PCC decompressor is disclosed, where compressed bitstream is received and decoded to obtain occupancy symbols, quantized positions using an octree decoding algorithm, obtains LoD of each of the points corresponding to the quantized positions, and obtain reconstructed prediction residuals, and reconstructs attributes respective of the reconstructed prediction residuals using a prediction algorithm, such as interpolation or weighted average calculations; see Chou’100 Fig. 9, and [0107]-[0109], where point cloud data is hierarchically organized in levels in 3D space, with a WxWxW unit cube for a voxel being partitioned into eight smaller sub cubes W/2xW/2xW/2, which can be further partitioned into smaller sub cubes; where the combined teaching suggests that the 3D points at a sub cube position can be interpolated by taking a distance based weighted average of nearest neighbors of the point).
Regarding claim 8, please see the above rejection of claim 7. Yea, Chou’100, Cao, and Chou’120 disclose the point cloud decoding device according to claim 7, wherein, in a case where there are two or more interpolated values with respect to each fractional precision position, the interpolation unit uses an average value of the two or more interpolated values as the interpolated value (see Yea Fig. 1 and [0007]-[0011], where attribute values of a 3D point are predicted by using a linear interpolation process of distances of nearest neighbors of a current point, where for an exemplified 3D point P2, neighboring three points P0, P5, and P4 are suggested; see Yea Fig. 5 and [0051]-[0057], where a G-PCC decompressor is disclosed, where compressed bitstream is received and decoded to obtain occupancy symbols, quantized positions using an octree decoding algorithm, obtains LoD of each of the points corresponding to the quantized positions, and obtain reconstructed prediction residuals, and reconstructs attributes respective of the reconstructed prediction residuals using a prediction algorithm, such as interpolation or weighted average calculations; see Chou’100 Fig. 9, and [0107]-[0109], where point cloud data is hierarchically organized in levels in 3D space, with a WxWxW unit cube for a voxel being partitioned into eight smaller sub cubes W/2xW/2xW/2, which can be further partitioned into smaller sub cubes; where the combined teaching suggests that the 3D points at a sub cube position can be interpolated by taking a distance based weighted average of nearest neighbors of the point, where for further partitioned sub cube positions, the neighboring points would be interpolated values and an average of nearest neighbors of interpolated values would be calculated for the further portioned sub cube point).
Regarding claim 9, please see the above rejection of claim 6. Yea, Chou’100, Cao, and Chou’120 disclose the point cloud decoding device according to claim 6, wherein the prediction circuitry uses, to generate the predicted value, an attribute value of a closest point with respect to each point in a frame of integer precision among points of the reference frame of fractional precision after motion compensation based on the motion information (see Yea Fig. 1, [0011]-[0014], and Table 1, where if variability is higher than a threshold, a rate distortion optimized predictor selection is performed where multiple predictor candidate or candidate predicted values are created based on a result of a neighbor point search in generating LoD, where attribute values of a nearest neighbor point is selected; see Yea Fig. 5 and [0051]-[0057], where a G-PCC decompressor is disclosed, where compressed bitstream is received and decoded to obtain occupancy symbols, quantized positions using an octree decoding algorithm, obtains LoD of each of the points corresponding to the quantized positions, and obtain reconstructed prediction residuals, and reconstructs attributes respective of the reconstructed prediction residuals using a prediction algorithm, such as interpolation or weighted average calculations; see Chou’100 Fig. 9, and [0107]-[0109], where point cloud data is hierarchically organized in levels in 3D space, with a WxWxW unit cube for a voxel being partitioned into eight smaller sub cubes W/2xW/2xW/2, which can be further partitioned into smaller sub cubes; see Chou’100 [0081]-[0082], where decoded motion data is used to produce motion compensated prediction values; where the combined teaching suggests that motion compensated attributes of integer unit cube positions may be used as a predicted attribute value for sub cube positions).
Regarding claim 10, please see the above rejection of claim 6. Yea, Chou’100, Cao, and Chou’120 disclose the point cloud decoding device according to claim 6, wherein the fractional precision is 1/2 voxel precision (see Chou’100 Fig. 3, [0046], [0049], [0054]-[0055], Fig. 9, and [0107]-[0109], where the point cloud data is hierarchically organized in levels in 3D space, with a WxWxW unit cube for a voxel being partitioned into eight smaller sub cubes W/2xW/2xW/2, which can be further partitioned into smaller sub cubes).
Regarding claim 11, it recites a method performing the device functions of claim 1. Yea, Chou’100, Cao, and Chou’120 teach a method performing the device functions of claim 1. Please see above for detailed claim analysis.
Please see the above rejection for claim 1, as the rationale to combine the teachings of Yea, Chou’100, Cao, and Chou’120 are similar, mutatis mutandis.
Regarding claim 12, it recites a non-transitory computer-readable medium including a program for causing a computer to perform the device functions of claim 1. Yea, Chou’100, Cao, and Chou’120 teach a program for causing a computer to perform the device functions of claim 1 (see Yea [0095], [0109]-[0110], [0119]-[0122], where the disclosed teachings are implemented by processors executing programs stored in non-transitory computer readable mediums). Please see above for detailed claim analysis.
Please see the above rejection for claim 1, as the rationale to combine the teachings of Yea, Chou’100, Cao, and Chou’120 are similar, mutatis mutandis.
Regarding claim 13, it recites a method performing the device functions of claim 6. Yea, Chou’100, Cao, and Chou’120 teach a method performing the device functions of claim 6. Please see above for detailed claim analysis.
Please see the above rejection for claim 6, as the rationale to combine the teachings of Yea, Chou’100, Cao, and Chou’120 are similar, mutatis mutandis.
Regarding claim 14, it recites a non-transitory computer-readable medium including a program for causing a computer to perform the device functions of claim 6. Yea, Chou’100, Cao, and Chou’120 teach a program for causing a computer to perform the device functions of claim 6 (see Yea [0095], [0109]-[0110], [0119]-[0122], where the disclosed teachings are implemented by processors executing programs stored in non-transitory computer readable mediums). Please see above for detailed claim analysis.
Please see the above rejection for claim 6, as the rationale to combine the teachings of Yea, Chou’100, Cao, and Chou’120 are similar, mutatis mutandis.
Regarding claim 15, Yea, Chou’100, Cao, and Chou’120 disclose a point cloud coding device comprising:
interpolation circuitry (see Yea [0095], [0109]-[0110], [0119]-[0122], where the disclosed teachings are implemented by processors executing programs stored in non-transitory computer readable mediums) configured to perform an interpolation process on a reconstructed point cloud of a coded frame with respect to geometry and attribute information indicating a geometry and an attribute of a point cloud, and generate reference frame of fractional-precision (see Yea Fig. 1 and [0007]-[0011], where in graph based point cloud compression (G-PCC), levels of detail (LoD) of each 3D point are generated based on a distance of each 3D point and attributes values of 3D points in each LoD is encoded by applying prediction in an LoD based order, where if a variability of a neighborhood of a 3D point is lower than a threshold, the distance based weighted average prediction is conducted by predicting attribute values using a linear interpolation process based on distances of nearest neighbors of current point; see Yea Fig. 4 and [0037]-[0047], where the G-PCC compressor receives positions of points in an input point cloud, an octree encoder encodes filtered and quantized positions into occupancy symbols of an octree representing the input point cloud using an octree encoding algorithm and produces reordered positions, a prediction module obtains prediction residuals respectively of received attributes of points in the input point cloud by applying a prediction algorithm to the received attributes in an order based on received LoD of each of the points corresponding to reordered positions from the octree encoder, where the prediction algorithm may include interpolation, weighted average calculation, a nearest neighbor algorithm and RDO; see Chou’100 Fig. 3, [0046], [0049], [0054]-[0055], Fig. 9, and [0107]-[0109], where received point cloud data includes geometry data for points as well as attributes of occupied points, and an octree coder compresses the geometry data using a octree scanning of points in 3D space, where the point cloud data is hierarchically organized in levels in 3D space, with a WxWxW unit cube for a voxel being partitioned into eight smaller sub cubes W/2xW/2xW/2, which can be further partitioned into smaller sub cubes), wherein the interpolation circuitry determines neighboring points among integer-precision voxels when a distance between coordinates of two voxels is below a threshold (p), and performs the interpolation process only between such neighboring voxels to generate fractional-precision geometry and attribute values (see Yea Fig. 4 and [0046]-[0047], where the prediction algorithm may include interpolation, weighted average calculation, a nearest neighbor algorithm, and RDO; see Yea [0064]-[0073], where nearest neighbor search is performed for each point of the point cloud, and can be conducted both at the encoder and the decoder; see Chou’120 [0160], where a threshold is used upon the distance between voxels i and j to control the size of the neighborhood of a voxel for filtering geometry elements to smooth the surface; where the combined teaching suggests that a neighborhood size can be controlled by comparing distances between received attribute points when performing the neighborhood search for applying the prediction algorithm);
motion estimation circuitry configured to perform motion estimation between the reference frame of fractional precision and a frame of integer precision to generate motion information (see Yea [0060]-[0063], where a geometry based or joint geometry/attribute based global/local motion estimation may be performed for motion compensation when determining prediction values; see Chou’100 [0060]-[0061], where estimating the motion of attributes of the block with respect to one or more reference frames of point cloud data; see Cao Fig. 4, [0083]-[0084], Fig. 6, and [0088]-[0089], where global and local motion estimation is performed and applied to perform motion compensation and encode motion vector information and point information, where local node motion estimation is applied to estimate a local motion at a finer scale and a node level in an octree, where local node motion vector of a node of the octree is a motion vector that is only applied on points within the node in prediction (reference) frame, and a G-PCC encoder may determine the local node motion vector in a recursive manner using a cost function to choose the best suitable motion vector, and assigning each sub-node a respective motion vector); and prediction circuitry configured to generate a predicted value based on the motion information (see Yea [0060]-[0063], where a geometry based or joint geometry/attribute based global/local motion estimation may be performed for motion compensation when determining prediction values; where the combined teachings suggests to perform global and local motion estimation for motion compensation, where local node motion estimation is performed to estimate a local motion at a finer scale and a node level in an octree by determining local node motion vectors of a node of the octree that is only applied on points within the node in reference frame);
entropy coding circuitry configured to entropy-code a difference between a point cloud of the coded frame and the predicted value (see Chou’120 [0063]-[0068], where motion data is estimated and provided for entropy encoding and inter-frame compression can be determined for compressing a given block, where encodes and transmits the differences between prediction values and corresponding original attributes, where the differences provide values of the residuals, and values of the prediction residual are encoded using region-adaptive hierarchical transformer, quantizer, and entropy encoder; see Yea [0049], where prediction residuals are quantized and entropy encoding is performed to obtain a compressed bitstream; see also Chou’100 [0058], Cao [0060] and [0064); where the combined teaching suggests to apply inter-frame compression when performing motion compensated encoding of point cloud data, where differences between prediction values and corresponding original attributes of the point cloud data are determined, quantized and entropy encoded to obtain and output compressed bitstream).
Please see the above rejection for claim 1, as the rationale to combine the teachings of Yea, Chou’100, Cao, and Chou’120 are similar, mutatis mutandis.
Regarding claim 16, Yea, Chou’100, Cao, and Chou’120 disclose a point cloud decoding device comprising:
interpolation circuitry (see Yea [0095], [0109]-[0110], [0119]-[0122], where the disclosed teachings are implemented by processors executing programs stored in non-transitory computer readable mediums) configured to perform an interpolation process on a reconstructed point cloud of a decoded frame with respect to geometry and attribute information indicating a geometry and an attribute of a point cloud, and generate a reference frame of fractional-precision (see Yea Fig. 1 and [0007]-[0011], where in graph based point cloud compression (G-PCC), levels of detail (LoD) of each 3D point are generated based on a distance of each 3D point and attributes values of 3D points in each LoD is encoded by applying prediction in an LoD based order, where if a variability of a neighborhood of a 3D point is lower than a threshold, the distance based weighted average prediction is conducted by predicting attribute values using a linear interpolation process based on distances of nearest neighbors of current point; see Yea Fig. 5 and [0051]-[0057], where a G-PCC decompressor is disclosed, where compressed bitstream is received and decoded to obtain occupancy symbols, quantized positions using an octree decoding algorithm, obtains LoD of each of the points corresponding to the quantized positions, and obtain reconstructed prediction residuals, and reconstructs attributes respective of the reconstructed prediction residuals using a prediction algorithm, such as interpolation or weighted average calculations; see Chou’100 Fig. 4 and [0076], Fig. 9, and [0107]-[0109], where compressed geometry data is received and decompressed, where the point cloud data is hierarchically organized in levels in 3D space, with a WxWxW unit cube for a voxel being partitioned into eight smaller sub cubes W/2xW/2xW/2, which can be further partitioned into smaller sub cubes), wherein the interpolation circuitry determines neighboring points among integer-precision voxels when a distance between coordinates of two voxels is below a threshold (p), and performs the interpolation process only between such neighboring voxels to generate fractional-precision geometry and attribute values (see Yea Fig. 4 and [0057], where the prediction algorithm may include interpolation, weighted average calculation, a nearest neighbor algorithm, and RDO; see Yea [0064]-[0073], where nearest neighbor search is performed for each point of the point cloud, and can be conducted both at the encoder and the decoder; see Chou’120 [0160], where a threshold is used upon the distance between voxels i and j to control the size of the neighborhood of a voxel for filtering geometry elements to smooth the surface; where the combined teaching suggests that a neighborhood size can be controlled by comparing distances between received attribute points when performing the neighborhood search for applying the prediction algorithm);
entropy decoding circuitry configured to decode motion information and a predicted residual from a bit stream (see Yea [0052], where compressed bitstream is received and decoded using various entropy decoding algorithms to obtain occupancy symbols and quantized prediction residuals; see Chou’120 [0077]-[0079] and [0084], where entropy decoding is performed to decode quantized transform coefficients from received encoded data, and process residual values for any of the attributes of occupied points and decode motion data);
prediction circuitry configured to generate a predicted value based on the motion information and the reference frame of fractional precision (see Yea [0057], where attributes are reconstructed respective of reconstructed prediction residuals using a prediction algorithm, such as interpolation or weighted average calculations; see Yea [0060]-[0063], where a geometry based or joint geometry/attribute based global/local motion estimation may be performed for motion compensation when determining prediction values; see Chou’100 [0081]-[0082], where decoded motion data is used to produce motion compensated prediction values; see Cao [0083], where a G-PCC coder may apply local node motion compensation only on a portion of the reference frame, where local node motion vector of a node of the octree is a motion vector that is only applied on points within the node in prediction (reference) frame); and
attribute information decoding circuitry configured to use a sum of the predicted residual and the predicted value as a decoding value of an attribute value (see Chou’120 [0086], where reconstructed residual values are combined with the prediction values to produce a reconstruction of the attributes of occupied points for the current point cloud frame).
Please see the above rejection for claim 1 and 15, as claim 16 is a corresponding decoding device to the encoding device of claim 15, and the rationale to combine the teachings of Yea, Chou’100, Cao, and Chou’120 are similar, mutatis mutandis.
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to TIMOTHY WING HO CHOI whose telephone number is (571)270-3814. The examiner can normally be reached 9:00 AM to 5:00 PM.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, VINCENT RUDOLPH can be reached at (571) 272-8243. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/TIMOTHY CHOI/Examiner, Art Unit 2671
/VINCENT RUDOLPH/Supervisory Patent Examiner, Art Unit 2671
Prosecution Insights