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 Arguments
Applicant’s arguments with respect to claim(s) 18-27, 31-34 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Claim Rejections - 35 USC § 103
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.
Claim(s) 18, 31-33 is/are rejected under 35 U.S.C. 103 as being unpatentable over Ma et al. (CN109063723) in view of Liu et al. (US2022/0044407).
To claim 18, Ma teach a method for generating a segmentation mask (annotation) of at least one input image, the method comprising:
providing a semantic segmentation model with a reference data set of reference superpixels wherein each reference superpixel is associated with a class (paragraph 0014, training images and their corresponding class labels constitute the training dataset);
generating a plurality of superpixels for the at least one image (paragraphs 0024, 0073, 0088, superpixel segmentation);
automatically generating labels for the generated plurality of the superpixels by generating said labels by identifying a most similar reference superpixel from the reference data set of reference superpixels (paragraph 0093, labeling superpixel regions; paragraphs 0017, 0064, multi-label classification network selects the object class corresponding to the highest probability as the classification result corresponding to the input image);
automatically generating a segmentation mask resulting in a masked image with the labeled superpixels (obvious in paragraphs 0024, 0073, where superpixel region is identified through iterative training);
automatically checking and correcting the segmentation mask for logical constrains to generate a corrected masked image (obvious in paragraphs 0038-0042, iteratively training);and
computer-aided checking of the corrected masked image for correctness, by labeling unlabeled superpixels and correcting labels of incorrectly labeled superpixels by assigning the incorrectly labeled superpixels to the correct class to generate a final masked image (paragraphs 0006, 0032, 0041-0042, 0096-0097, re-identifying superpixel regions to obtain updated superpixel regions and corresponding new region labels, correcting the erroneous labels); and
training the semantic segmentation model by adding all of the labeled superpixels of the final masked image to the reference data set according to the class associated with each labeled superpixel of the final masked image (paragraphs 0098-0105, updated region labels of superpixel regions would be those of newly labeled and label corrected, to improve accuracy of the semantic segmentation model by iterative training).
In furthering said obviousness, Liu teach a method for generating a segmentation mask of at least one input image, the method comprising: automatically generating a segmentation mask resulting in a masked image with the labeled superpixels (paragraphs 0090-0095, input labeled image to generate segmentation mask); automatically checking and correcting the segmentation mask for logical constraints to generate a corrected masked image (paragraphs 0094-0102, refine segmentation mask).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate teaching of Liu into the method of Ma, in order to further implementation of segmentation mask generation and refinement.
To claim 31, Ma and Liu teach a system for generating a segmentation mask of at least one image (as explained in response to claim 18 above).
To claim 32, Ma and Liu teach a computer-readable non-transitory storage medium, comprising a computer program having instructions for carrying out the method of claim 18 (as explained in response to claim 18 above).
To claim 33, Ma and Liu teach a system for monitoring a condition of a tool of a machining tool (as explained in response to claim 18 above).
Claim(s) 19-20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Ma et al. (CN109063723) in view of Liu et al. (US2022/0044407) and Jampani et al. (US2019/0340728).
To claim 19, Ma and Liu teach claim 18.
Ma teach wherein the generating a plurality of the superpixels is done by using a Superpixel Sampling Network (despite of terminology, obvious in paragraphs 0024-0025).
In further said obviousness, Jampani teach generating a plurality of the superpixels is done by using a Superpixel Sampling Network (abstract), which would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate into the method of Ma and Liu, in order to obtain superpixel regions.
To claim 20, Ma, Liu and Jampani teach claim 19.
Ma, Liu and Jampani teach wherein the Superpixel Sampling Network is based on a combination of a Simple Linear Iterative Clustering algorithm and a deep neural network (Ma, paragraph 0024; Jampani, Fig. 2, paragraph 0033).
Claim(s) 21-23 is/are rejected under 35 U.S.C. 103 as being unpatentable over Ma et al. (CN109063723) in view of Liu et al. (US2022/0044407) and Price et al. (US2015/0206315).
To claim 21, Ma and Liu teach claim 18.
Ma and Liu teach wherein the automatically generating labels comprises calculating a similarity of the superpixels with the reference superpixels from the reference data set (Ma, obvious in paragraphs 0017, 0064, probability representation of similarity with reference/training data)
In further said obviousness, Price teach automatically generating labels comprises calculating a similarity of the superpixels with the reference superpixels from the reference data set (paragraphs 0025, 0030), which would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate into the method of Ma and Liu, in order to further implementation of labeling.
To claim 22, Ma, Liu and Price teach claim 21.
Ma, Liu and Price teach wherein the similarity of a superpixel of the plurality of the superpixels with each reference superpixel from the reference data set is calculated and this calculation is repeated for each superpixel of the plurality of the superpixels (Ma, paragraphs 0035-0047, iterative operation).
To claim 23, Ma, Liu and Price teach claim 21.
Ma, Liu and Price teach wherein the calculating the similarity comprises setting a similarity threshold (Ma, paragraph 0080).
To claim 27, Ma, Liu and Price teach claim 21.
Ma, Liu and Price teach wherein the reference data set comprises one or more subsets, wherein each subset comprises reference superpixels associated to one class of the classes and different subsets comprise reference superpixels associated with different classes (Ma, paragraphs 0014, 0060).
Claim(s) 25-26 is/are rejected under 35 U.S.C. 103 as being unpatentable over Ma et al. (CN109063723) in view of Liu et al. (US2022/0044407), Price et al. (US2015/0206315) and Sang et al. (CN106203255).
To claim 25, Ma, Liu and Price teach claim 21.
Though Ma, Liu and Price do not expressly disclose wherein the calculating the similarity comprises applying a Distance Metric Learning method to the superpixels and to the reference superpixels, such feature is well-known practice in the art.
Sang teach pedestrian re-identification wherein calculating the similarity comprises applying a Distance Metric Learning method to the superpixels and to the reference superpixels (paragraphs 0008-0069, metric learning used in pedestrian recognition, often called person re-identification or Re-ID, is a direct application of distance metric learning), which would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate into the method of Ma, Liu and Price, in order to further implementation of identification.
To claim 26, Ma, Liu, Price and Sang teach claim 25.
Ma, Liu, Price and Sang teach wherein the Distance Metric Learning method comprises extracting features from superpixel images with a feature extraction module, wherein the feature extraction module embeds the extracted superpixel image features into a feature space so that the more related the superpixel images are to each other, the closer encoding vectors are in the feature space, and defining a similarity of two superpixel images by a distance of their vectors in the feature space (as explained in response to claim 25 above; it is also well-known in the art that the core goal of metric learning or distance metric learning is to learn a task-specific feature space where the distance between data points reflects their actual semantic similarity; and task such as pedestrian recognition of Sang would involve converting a complex visual relationship into a simple mathematical distance, wherein a smaller distance directly equals higher similarity).
Claim(s) 24 is/are rejected under 35 U.S.C. 103 as being unpatentable over Ma et al. (CN109063723) in view of Liu et al. (US2022/0044407) and Price et al. (US2015/0206315) and Liao (US2015/0302275).
To claim 24, Ma, Liu and Price teach claim 23.
Ma, Liu and Price do not expressly disclose wherein, when the calculated similarity is below the set similarity threshold, a notification is issued.
Liao teach sending notification to prompt user whether to continue processing when calculated similarity is below the set similarity threshold (paragraphs 0021, 0036), which would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate into the method of Ma, Liu and Price, in order to notify user operation process.
Claim(s) 34 is/are rejected under 35 U.S.C. 103 as being unpatentable over Ma et al. (CN109063723) in view of Liu et al. (US2022/0044407) and Zhou et al. (CN108447050).
To claim 34, Ma and Liu teach claim 18.
But, Ma and Liu do not expressly disclose wherein the at least one input image is an image of a tool of a machining tool and the trained semantic segmentation model performs analysis of the images of the tool to infer the condition of the tool.
Zhou teach using superpixel in image analysis, wherein at least one input image is an image of a tool of a machining tool and the trained segmentation model performs analysis of the images of the tool to infer the condition of the tool (Fig. 2; paragraphs 0004, 0008-0020), which would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate into the method of Ma, Liu and Price, in order to further image analysis application by design preference.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). 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 ZHIYU LU whose telephone number is (571)272-2837. The examiner can normally be reached Weekdays: 8:30AM - 5:00PM.
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, Stephen R Koziol can be reached at (408) 918-7630. 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.
ZHIYU . LU
Primary Examiner
Art Unit 2669
/ZHIYU LU/Primary Examiner, Art Unit 2665 October 3, 2026
Prosecution Insights