CTNF 18/846,639 CTNF 82525 Notice of Pre-AIA or AIA Status 07-03-aia AIA 15-10-aia The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA. Claim Rejections - 35 USC § 112 07-30-02 AIA The following is a quotation of 35 U.S.C. 112(b): (b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention. The following is a quotation of 35 U.S.C. 112 (pre-AIA), second paragraph: The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention. Claims 1 and 17 recites the limitation "said reconstructed point cloud” in lines 11 respectively. There is insufficient antecedent basis for this limitation in the claims. Claims 2-4, 6, 8-14 and 18-24 depends directly or indirectly on claims 1 and 11, therefore they are rejected. Claim Rejections - 35 USC § 103 07-06 AIA 15-10-15 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. 07-20-aia AIA The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action: A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made. 07-21-aia AIA Claim s 1-4, 6, 8-12 and 17-22 are rejected under 35 U.S.C. 103 as being unpatentable over NPL1 (Self-Supervised Deep Learning on Point Clouds by Reconstructing Space, Jonathan Sauder et al., arXiv, 2019, Pages 1-11) hereafter NPL1 (single reference 103 as the claim limitations are show/discloses in multiple figs) . 1. Regarding claim 1 as best understood by the examiner, NPL1 discloses a method (Page 4 section 3 discloses a method) , comprising: partitioning an input point cloud into a plurality of chunks by a first neural network-based module (Page 2 fig 1 also shows and discloses “the original object is split into voxels” and page 4 section 3 discloses “In this paper we propose a self-supervised method that learns powerful representations from raw point cloud data. Our method works by training a neural network (i.e first neural network) to reassemble point clouds whose parts have been randomly displaced. The key assumption of the proposed method is that learning to reassemble displaced point cloud segments is only possible by learning holistic representations that capture the high-level semantics of the objects in the point cloud . We phrase the self-supervised learning task as a point segmentation task (partitioning), in which the label for each point (i.e chunks) is generated from the point cloud itself with the following procedure: the input point cloud (i.e input point cloud) is scaled to unit cube before each axis is split into k equal lengths, forming k3 voxels. We use these to assign each point its voxel ID as a label. Subsequently all voxels are randomly swapped with other voxels and a neural network is trained to predict the original voxel ID of each point.” meeting the above claim limitations, examiner notes that the specifics of “partitioning, chunks and input point cloud and “a first neural network” are not required by the current claim) ; for each chunk of points from said input point cloud: generating, by a second neural network-based module, a respective codeword that describes at least a shape in said point cloud chunk (Pages 2, 5-6 fig 1-3 also shows and discloses “the original object is split into voxels”, section C predicts the voxels labels (i.e code words) and page 5 tables 1, 3, section 4, 4.2 shows and discloses we train our model using “ShapeNet” meeting the limitations of describes at least the shape in the point cloud i.e points/ voxels/chunks, and page 4 section 3 discloses “In this paper we propose a self-supervised method that learns powerful representations from raw point cloud data. Our method works by training a neural network (i.e first neural network) to reassemble point clouds whose parts have been randomly displaced. The key assumption of the proposed method is that learning to reassemble displaced point cloud segments is only possible by learning holistic representations that capture the high-level semantics of the objects in the point cloud. We phrase the self-supervised learning task as a point segmentation task (partitioning), in which the label for each point (i.e each chunk) is generated from the point cloud itself with the following procedure: the input point cloud (i.e input point cloud) is scaled to unit cube before each axis is split into k equal lengths, forming k3 voxels. We use these to assign each point its voxel ID as a label. Subsequently all voxels are randomly swapped with other voxels and a neural network (i.e second neural network) is trained to predict the original voxel ID of each point (i.e codeword) .” meeting the above claim limitations, examiner notes that the specifics of “a second neural network” are not required by the current claim) ; and reconstructing, by a third neural network-based module, said point cloud chunk based on said codeword to form a respective reconstructed point cloud chunk (pages 5- 6, figs 2-3 shows and discloses “Figure 2a shows that a decrease in self-supervised training loss on ShapeNet gives a better downstream classification accuracy on ModelNet40 (i.e by third neural network) , which suggests that correctly reconstructing the point cloud parts results requires learning representations that capture the semantics of the objects at hand. The obtained embeddings from a DGCNN without method for the ModelNet10 test data are visualized using t-SNE [17] in Figure 2b. One can see that clear, separable clusters are formed for each class except for the classes dresser (violet) vs nightstand (pink), which are almost visually indiscernible when scaled to unit cube, as done in the ShapeNet dataset” meeting the above claim limitations) ; reconstructing said input point cloud based on said respective reconstructed point cloud chunks (pages 1-2, 5-6 and figs 1(d)-3 and abstract shows and discloses reconstructing said input cloud based on said reconstructed input point cloud chunks (i.e points/voxels) meeting the claim limitations) ; obtaining a mismatch metric based on said reconstructed point cloud and said input point cloud (fig 1 d shows the misclassifications in (Red) (i.e mismatch metric) meeting the above claim limitations, examiner notes that the specifics of “mismatch metric” are not required by the current claim) ; and adjusting parameters of said first neural network-based module, said second neural network-based module, and said third neural network-based module, based on said mismatch metric (figs 2-3, pages 3 section 2.2,pages 5-7 shows and discloses “Training epochs” and fig 3 shows the reduction in the training loss (i.e loss is reduced based on the adjustment of the parameters of the models) i.e adjusting parameters for each of the first, second and third neural networks based on the misclassifications in (red) i.e mismatch metric in-order to increase the accuracy) . Before the effective filing date of the invention was made, different figs in NPL1 are combinable. The suggestion/motivation would be an improved, efficient and accurate method/system on page 8 section 6. 2. Regarding claim 2, NPL1 discloses the method of claim 1, wherein said point cloud chunk is a part of said input point cloud located at a random position with a random size in said input point cloud (fig 1 (b) shows the voxels (chunks) are randomly arranged meeting the above claim limitations) . 3. Regarding claim 3, NPL1 discloses the method of claim 1, wherein said first neural network-based module corresponds to a PointNet (section 4.2 (segmentation) task shows “PointNet” with 89.2 accuracy meeting the claim limitations) . 4. Regarding claim 4, NPL1 discloses the method of claim 1, wherein said second neural network-based module corresponds to a FoldingNet (page 5 Table 1 shows Foldingnet accuracy 94.40% meeting the claim limitations) . 5. Regarding claim 6, NPL1 discloses the method of claim 1, wherein said third neural network- based module corresponds to PointNet++ or VoteNet (page 6 section 4.3 shows PointNet++ with accuracy 90.7% meeting the claim limitations) . 6. Regarding claim 8, NPL1 discloses the method of claim 1, further comprising: identifying a set of codewords from said respective codewords; grouping said set of codewords into one or more clusters; and obtaining a representative codeword for each cluster of said one or more clusters (Page 4 section 3, fig 5 shows identifying a set of codewords from said respective codewords; grouping said set of codewords into one or more clusters; and obtaining a representative codeword for each cluster of said one or more clusters) . 7. Regarding claim 9, NPL1 discloses the method of claim 8, wherein said set of codewords are identified based on reconstruction quality of said plurality of reconstructed point cloud chunks (page 4 section 3, page 8 section 5 discloses wherein said set of codewords are identified based on reconstruction quality of said plurality of reconstructed point cloud chunks) . 8. Regarding claim 10, NPL1 discloses the method of claim 8, further comprising: reconstructing a primitive for a corresponding representative codeword, using said second neural network-based module (figs 2, 4-5 shows and discloses reconstructing a primitive for a corresponding representative codeword, using said second neural network-based module (i.e ShapeNet)) . 9. Regarding claim 11, NPL1 discloses the method of claim 1, further comprising: performing classification on another point cloud based on said first and third neural network-based modules (page 8 section 5 discloses downstream object classification task on the randomly drawn point cloud using the method (i.e based on the first and third network) meeting the claim limitations) . 10. Regarding claim 12, NPL1 discloses the method of claim 1, further comprising: performing object detection on another point cloud based on said first and third neural network-based modules (page 8 section 5 discloses downstream object classification task on the randomly drawn point cloud (i.e another point cloud) using the method (i.e based on the first and third network) meeting the claim limitations) . 11. Claim 17 is a corresponding apparatus claim of claim 1. See the corresponding explanation of claim 1. Examiner notes that an apparatus, comprising one or more processors and at least one memory coupled to said one or more processors, wherein said one or more processors are configured to perform the steps recited in claim 17 would be obvious and within one of ordinary skill in the art. Examiner notes that a computer (i.e one or more processors and the memory coupled to the memory (i.e one or more processors are configured to) would be obvious in view of Page 4 section 3 “Algorithm” disclosed by NPL1 . 12. Claim 18 is a corresponding apparatus claim of claim 8. See the corresponding explanation of claim 8. 13. Claim 19 is a corresponding apparatus claim of claim 9. See the corresponding explanation of claim 9. 14. Claim 20 is a corresponding apparatus claim of claim 10. See the corresponding explanation of claim 10. 15. Claim 21 is a corresponding apparatus claim of claim 11. See the corresponding explanation of claim 11. 16. Claim 22 is a corresponding apparatus claim of claim 12. See the corresponding explanation of claim 12 . 07-21-aia AIA Claim s 13-14 and 23-24 are rejected under 35 U.S.C. 103 as being unpatentable over NPL1 in view of NPL10 (Real-time Compression of Point Cloud Streams, Julius Kammerl et al., IEEE, 2012, Pages 778-785) hereafter NPL10 . 17. Regarding claim 13, NPL1 discloses the method of claim 1. NPL1 discloses further comprising: partitioning another point cloud into another plurality of point cloud chunks based on said first neural network-based module (Fig 1 and also section 3, page 8 section 5 discloses downstream object classification task on the randomly drawn point cloud (i.e another plurality of point cloud chunks) using the method (i.e partitioning another point cloud into another plurality of point cloud chunks based on said first neural network-based module) meeting the claim limitations) . NPL1 also discloses Autoencoders (i.e encoding) page 3 section 2.2. NPL1 however is silent and fails to disclose and encoding each of said plurality of point cloud chunks into one or more bitstream. NPL10 discloses encoding each of said plurality of point cloud chunks into one or more bitstream (Figs 1,3-4, 5, 7-8 and pages 778-780, 782-783 shows and discloses encoding (compression) each of said plurality of point cloud chunks (voxels) into one or more bitstream (i.e stream)) . Before the effective filing date of the invention was made, NPL1 and NPL10 are combinable because they are from the same filed of endeavor and are analogous art of point cloud data processing. The suggestion/motivation would be an increased point precision, reduced computation and memory requirements system/method (page 785 section VIII) . Therefore, it would be obvious and within one of ordinary skill in the art to have recognized the advantages of NPL10 in the method/system of NPL1 to obtain the invention as specified in claim 13. 18. Regarding claim 14, NPL1 discloses the method of claim 1. NPL1 discloses the reconstruction of the point cloud chunks and the autoencoder (Fig 1 and also section 3, page 8 section 5 discloses downstream object classification task on the randomly drawn point cloud (i.e another plurality of point cloud chunks) using the method (i.e partitioning another point cloud into another plurality of point cloud chunks based on said first neural network-based module) meeting the claim limitations) . NPL1 is silent and however fails to disclose further comprising: decoding one or more bitstreams to form another plurality of point cloud chunks; and reconstructing another point cloud from said another plurality of point cloud chunks. NPL10 discloses decoding one or more bitstreams to form another plurality of point cloud chunks; and reconstructing another point cloud from said another plurality of point cloud chunks (figs 1, 7, pages 781-782 shows and discloses decompression of the encoded bitstreams and reconstruction of the point cloud components (as seen in fig 7 decoding side) to get the final point clout output meeting the above claim limitations) . Before the effective filing date of the invention was made, NPL1 and NPL10 are combinable because they are from the same filed of endeavor and are analogous art of point cloud data processing. The suggestion/motivation would be an increased point precision, reduced computation and memory requirements system/method (page 785 section VIII) . Therefore, it would be obvious and within one of ordinary skill in the art to have recognized the advantages of NPL10 in the method/system of NPL1 to obtain the invention as specified in claim 14. 19. Claim 23 is a corresponding apparatus claim of claim 13. See the corresponding explanation of claim 13. 20. Claim 24 is a corresponding apparatus claim of claim 14. See the corresponding explanation of claim 14. Examiner's Note: Examiner has cited figures, and paragraphs in the references as applied to the claims above for the convenience of the applicant. Although the specified citations are representative of the teachings in the art and are applied to the specific limitations within the individual claim, other passages and figures may apply as well. It is respectfully requested for the applicant, in preparing the responses, to fully consider the references in entirety as potentially teaching all or part of the claimed invention, as well as the context of the passage as taught by the prior art or disclosed by the examiner. Examiner has also cited references in PTO892 but not relied on, which are relevant and pertinent to the applicant’s disclosure, and may also be reading (anticipatory/obvious) on the claims and claimed limitations. Applicant is advised to consider the references in preparing the response/amendments in-order to expedite the prosecution. Conclusion Any inquiry concerning this communication or earlier communications from the examiner should be directed to JAYESH PATEL whose telephone number is (571)270-1227. The examiner can normally be reached IFW Mon-FRI. 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, Andrew Bee can be reached at 571-270-5183. 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. /JAYESH PATEL/ Primary Examiner Art Unit 2677 /JAYESH A PATEL/Primary Examiner, Art Unit 2677 Application/Control Number: 18/846,639 Page 2 Art Unit: 2677 Application/Control Number: 18/846,639 Page 3 Art Unit: 2677 Application/Control Number: 18/846,639 Page 4 Art Unit: 2677 Application/Control Number: 18/846,639 Page 5 Art Unit: 2677 Application/Control Number: 18/846,639 Page 6 Art Unit: 2677 Application/Control Number: 18/846,639 Page 7 Art Unit: 2677 Application/Control Number: 18/846,639 Page 8 Art Unit: 2677 Application/Control Number: 18/846,639 Page 9 Art Unit: 2677 Application/Control Number: 18/846,639 Page 10 Art Unit: 2677 Application/Control Number: 18/846,639 Page 11 Art Unit: 2677 Application/Control Number: 18/846,639 Page 12 Art Unit: 2677 Application/Control Number: 18/846,639 Page 13 Art Unit: 2677
Prosecution Insights