DETAILED ACTION
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claims 19-38 are pending.
Claims 1-18 are canceled.
Prior Art:
D1: US20110116708A1 Zhou et al
D2: US20150071541A1 Qutub et al
D3: US20040145593A1 Berkner et al (new)
Claim Rejections - 35 USC § 103
The following is a quotation of pre-AIA 35 U.S.C. 103(a) which forms the basis for all obviousness rejections set forth in this Office action:
(a) A patent may not be obtained though the invention is not identically disclosed or described as set forth in section 102 of this title, if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains. Patentability shall not be negatived by the manner in which the invention was made.
Claim(s) 19-38 is/are rejected under 35 U.S.C. 103 as being unpatentable over D1 in view of D2 and further in view of D3.
Regarding claim 19, D1 teaches a computer-implemented method comprising:
obtaining an image comprising a plurality of image segments;
(D1, "input image is received by block 100", [0019]; "Typical natural images contain multiple regions with each image region being a set of pixels grouped based on homogeneity", [0003]; obtaining an image and segmenting it into a plurality of image regions)
associating individual image segments of the plurality of image segments with information indicative of possible interpretations of the individual image segments;
(D1, "computes the assigning probabilities of each region belonging to each object class", [0021]; associating individual regions with probabilities of belonging to an object class, indicating possible interpretations)
generating a plurality of combinations of the plurality of image segments;
(D1, "generating multiple hypotheses for region label assignment", [0006]; "multiple hypotheses of region combination are explored and determined", [0028]; generating a plurality of combinations by creating multiple hypotheses of region combinations)
generating, for each combination of the plurality of combinations, a set of relationships between image segments of the combination to produce a plurality of sets of relationships; and
(D1, " It also encodes information such as shape (where its boundary is) and context (who are neighboring regions)", [0024]; D2, "determining the connectivity of the one or more objects to each other based at least in part on a graphical analysis of the one or more objects", [Abstract]; "The object location and adjacency can be translated to a connectivity graph", [0128]; D1 teaches encoding context based on neighboring regions, while D2 explicitly teaches generating sets of relationships in the form of connectivity graphs based on object locations and adjacency)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to incorporate the teachings of D2 into the system or method of D1 in order to explicitly model network connectivity and relationships across objects using graph-based analysis. The combination of D1 and D2 also teaches other enhanced capabilities.
The combination of D1 and D2 further teaches:
determining information indicative of a most likely interpretation of the image,
(D1, "Image parsing is a process to assign object labels to each region so that the most probable interpretation of the input image can be achieved.", [0025]; determining a most likely interpretation by assigning labels to achieve the most probable interpretation of the image)
wherein determining the information includes processing the possible interpretations based on the plurality of sets of relationships,
(D1, "optimizing the label assignment so that an overall assigning probability for all regions is maximized", [0007]; "takes into account not only regions themselves but also the contextual relationship among regions", [0023]; processing the possible interpretations by optimizing the label assignment based on the contextual relationships among the regions)
wherein for a particular combination, positioning information associated with the image segments of the particular combination is generated, wherein the set of relationships for the particular combination includes the generated positioning information, and
(D2, "The object location and adjacency can be translated to a connectivity graph", [0128]; the relationships including positioning information via object location translated to the graph)
The combination of D1 and D2 does not expressly disclose but D3 teaches:
wherein the positioning information encodes relative positions including one or more of left/right, up/down, top/bottom, or parallel/perpendicular.
(D3, "If relative positioning is desired, the outcome could be a pair (left/right, top/bottom).", [0136]; "positioning attributes (with respect to the areas) including anchor point of segment, relative positioning (e.g., top, bottom, left, right)", [0082]; D3 teaches encoding relative positioning of image segments as left/right and top/bottom pairs)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to incorporate D3's known positional-descriptor scheme into D1/D2's object graph in order to predictably enhances spatial-relationship representation. The combination of D1, D2 and D3 also teaches other enhanced capabilities.
Regarding claims 20 and 30, the combination of D1, D2 and D3 teaches its/their respective base claim(s).
The combination further teaches the method of claim 19, wherein an individual combination of the plurality of combinations includes at least a pair of image segments.
(D1, "two nodes are connected if their corresponding regions are adjacent.", [0032]; a combination may include at least a pair of image segments connected through adjacency)
Regarding claims 22 and 32, the combination of D1, D2 and D3 teaches its/their respective base claim(s).
The combination further teaches the method of claim 19, wherein the positioning information includes relative positioning between the image segments of the particular combination.
(D2, "Adjacency can be measured by both object-object contact and distance between object centroids", [0128]; obtaining relative positioning by measuring distance between object centroids)
Regarding claims 24 and 34, the combination of D1, D2 and D3 teaches its/their respective base claim(s).
The combination further teaches the method of claim 19, wherein determining the information indicative of the most likely interpretation of the image is based on a surviving candidate interpretation of the image from a plurality of candidate interpretations of the image.
(D1, "the top best path is taken", [0036]; determining the interpretation based on a surviving candidate by selecting the top best path)
Regarding claims 25 and 35, the combination of D1, D2 and D3 teaches its/their respective base claim(s).
The combination further teaches the method of claim 24, further comprising:
generating a textual representation associated with the surviving candidate interpretation.
(D1, "Image parsing is a process to assign object labels to each region so that the most probable interpretation of the input image can be achieved.", [0025]; "For example, p(Si=“Same”|si=“Car”) can be learnt by counting how many times in the training set a region with label “Car” with all surrounding regions having label “Car” also.", [0038]; generating a textual representation by outputting text-based object labels, such as the string "Car")
Regarding claims 26 and 36, the combination of D1, D2 and D3 teaches its/their respective base claim(s).
The combination further teaches the method of claim 24, wherein the surviving candidate interpretation of the image is determined, at least in part, based on respective quality levels of the plurality of candidate interpretations.
(D1, "re-scored by another multi-class linear SVM", [0041]; evaluating candidate interpretations based on quality levels via SVM re-scoring)
Regarding claims 27 and 37, the combination of D1, D2 and D3 teaches its/their respective base claim(s).
The combination further teaches the method of claim 19, further comprising:
partitioning the image into the plurality of image segments, wherein partitioning is based on pixel mapping.
(D1, "pixels are grouped into super-pixels based on similarities", [0028]; partitioning based on pixel mapping by grouping pixels into super-pixels)
Regarding claims 28 and 38, the combination of D1, D2 and D3 teaches its/their respective base claim(s).
The combination further teaches the method of claim 19, wherein at least a subset of combinations of the plurality of combinations is identified as sharing a common aspect, the common aspect including spatial proximity.
(D1, "Each region is a set of pixels grouped based on similarities such as adjacency, color and smoothness.", [0024]; combinations share a common aspect including spatial proximity via adjacency)
Regarding claim 29, D1 teaches a computer-implemented method comprising:
obtaining an image formed from a plurality of image segments,
(D1, "input image is received by block 100", [0019]; "Typical natural images contain multiple regions with each image region being a set of pixels grouped based on homogeneity", [0003]; obtaining an image and segmenting it into a plurality of image regions)
wherein individual image segments are associated with information indicative of a set of image segment identigens, and
(D1, "computes the assigning probabilities of each region belonging to each object class.", [0021]; "Image parsing is a process to assign object labels to each region", [0025]; associating individual image segments with probabilities and object labels, which represent a set of image segment identigens)
wherein an individual image segment identigen reflects a possible interpretation of an associated image segment;
(D1, "semantics (what is the probability of the region belonging to each object class)", [0003]; "Image parsing is a process to assign object labels to each region so that the most probable interpretation of the input image can be achieved.", [0025]; the assigned object labels/classes reflect a possible semantic interpretation of the associated image region/segment)
generating a plurality of combinations of the plurality of image segments;
(D1, "generating multiple hypotheses for region label assignment", [0006]; "multiple hypotheses of region combination are explored and determined", [0028]; generating a plurality of combinations by creating multiple hypotheses of region combinations)
generating, for each combination of the plurality of combinations, a set of relationship identigens informing relationships between image segments of the combination; and
(D1, "It also encodes information such as shape (where its boundary is) and context (who are neighboring regions).", [0024]; D2, "determining, by at least one of the one or more computing devices, the connectivity of the one or more objects to each other based at least in part on a graphical analysis of the one or more objects", [Claim 1]; "The object location and adjacency can be translated to a connectivity graph", [0128]; D1 teaches encoding context based on neighboring regions, while D2 explicitly teaches generating sets of relationships in the form of connectivity graphs that inform the relationship between objects/segments)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention was made to incorporate the teachings of D2 into the system or method of D1 in order to explicitly model network connectivity and relationships across objects using graph-based analysis. The combination of D1 and D2 also teaches other enhanced capabilities.
The combination of D1 and D2 further teaches:
determining information indicative of an entigen group that represents a most likely interpretation of the image, wherein determining the information includes processing the image segment identigens based on the relationship identigens,
(D1, "Image parsing is a process to assign object labels to each region so that the most probable interpretation of the input image can be achieved.", [0025]; "optimizing the label assignment so that an overall assigning probability for all regions is maximized", [0007]; "takes into account not only regions themselves but also the contextual relationship among regions", [0023]; processing the possible interpretations by optimizing the label assignment based on the contextual relationships among the regions to determine the most probable overall interpretation of the image)
wherein for a particular combination, positioning information associated with the image segments of the particular combination is generated, wherein the set of relationships for the particular combination includes the generated positioning information,
(D2, see comments on claim 19)
and wherein the positioning information encodes relative positions including one or more of left/right, up/down, top/bottom, or parallel/perpendicular.
(D3, see comments on claim 19)
Response to Arguments
Applicant's arguments filed on 8/26/2026 with respect to one or more of the pending claims have been fully considered but are moot in view of the new ground(s) of rejection.
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 JIANXUN YANG whose telephone number is (571)272-9874. The examiner can normally be reached on MON-FRI: 8AM-5PM Pacific Time.
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, Amandeep Saini can be reached on (571)272-3382. 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.
/JIANXUN YANG/
Primary Examiner, Art Unit 2662 10/1/2026
