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 .
This office correspondence is in response to the application number19/078595 filed on March 13, 2025.
Claims 1 – 20 are pending.
Authorization for Internet Communications
The examiner encourages Applicant to submit an authorization to communicate with the examiner via the Internet by making the following statement (from MPEP 502.03):
“Recognizing that Internet communications are not secure, I hereby authorize the USPTO to communicate with the undersigned and practitioners in accordance with 37 CFR 1.33 and 37 CFR 1.34 concerning any subject matter of this application by video conferencing, instant messaging, or electronic mail. I understand that a copy of these communications will be made of record in the application file.”
Please note that the above statement can only be submitted via Central Fax (not Examiner's Fax), Regular postal mail, or EFS Web using PTO/SB/439.
Priority
This application is claiming the benefit of prior-filed application 18/306853 (now U.S. Patent 12,301,650) which further claims benefit to prior-filed application 17/372948 (now U.S. Patent 11,729,252) which further claims benefit to prior-filed application 16/356765 (now U.S. Patent 11,064,011) which further claimed benefit to prior-filed application 15/083976 (now U.S. Patent 10270839) under 35 U.S.C. 120, 121, 365(c), or 386(c). Co-pendency between the current application and the prior application 18/306853 is required. Since the instant application and application 18/306853 were co-pending at the time of the filing date of the instant application, the applicant is entitled to the benefit claim to the prior-filed application, which is a priority date of 3/29/2016.
Information Disclosure Statement
The information disclosure statement(s) (IDS) submitted on 03/13/2025, 07/29/2025, 11/04/2025, 3/21/2026 and 07/09/2026 were filed after the mailing date of the application filing date on March 13, 2025. The submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statements are being considered by the examiner.
Double Patenting
The non-statutory double patenting rejection is based on a judicially created doctrine grounded in public policy (a policy reflected in the statute) so as to prevent the unjustified or improper timewise extension of the “right to exclude” granted by a patent and to prevent possible harassment by multiple assignees. A non-statutory double patenting rejection is appropriate where the conflicting claims are not identical, but at least one examined application claim is not patentably distinct from the reference claim(s) because the examined application claim is either anticipated by, or would have been obvious over, the reference claim(s). See, e.g., In re Berg, 140 F.3d 1428, 46 USPQ2d 1226 (Fed. Cir. 1998); In re Goodman, 11 F.3d 1046, 29 USPQ2d 2010 (Fed. Cir. 1993); In re Longi, 759 F.2d 887, 225 USPQ 645 (Fed. Cir. 1985); In re Van Ornum, 686 F.2d 937, 214 USPQ 761 (CCPA 1982); In re Vogel, 422 F.2d 438, 164 USPQ 619 (CCPA 1970); In re Thorington, 418 F.2d 528, 163 USPQ 644 (CCPA 1969).
A timely filed terminal disclaimer in compliance with 37 CFR 1.321(c) or 1.321(d) may be used to overcome an actual or provisional rejection based on non-statutory double patenting provided the reference application or patent either is shown to be commonly owned with the examined application, or claims an invention made as a result of activities undertaken within the scope of a joint research agreement. See MPEP § 717.02 for applications subject to examination under the first inventor to file provisions of the AIA as explained in MPEP § 2159. See MPEP §§ 706.02(l)(1) - 706.02(l)(3) for applications not subject to examination under the first inventor to file provisions of the AIA . A terminal disclaimer must be signed in compliance with 37 CFR 1.321(b).
The USPTO Internet website contains terminal disclaimer forms which may be used. Please visit www.uspto.gov/patent/patents-forms. The filing date of the application in which the form is filed determines what form (e.g., PTO/SB/25, PTO/SB/26, PTO/AIA /25, or PTO/AIA /26) should be used. A web-based eTerminal Disclaimer may be filled out completely online using web-screens. An eTerminal Disclaimer that meets all requirements is auto-processed and approved immediately upon submission. For more information about eTerminal Disclaimers, refer to www.uspto.gov/patents/process/file/efs/guidance/eTD-info-I.jsp.
Claims 11 is rejected on the ground of non-provisional non-statutory anticipatory-type double patenting as being unpatentable over claim 1 of U.S. Patent 12,301,650. Although the conflicting claims are not identical, they are not patently distinct from each other because both sets of claims are directed to the same invention. This is a non-provisional non-statutory anticipatory-type double patenting rejection since the claims directed to the same invention have in fact been patented.
In regard to claim 11:
11. A method for automated content curation, comprising:
1. A method for network communication of image based content collections, the method comprising:
receiving content messages from a plurality of client devices;
storing, at a database of a server computer system, a plurality of content messages from a plurality of client devices including a first mobile device and a second mobile device;
analyzing the content messages using computer vision to determine image quality metrics;
generating, by the server computer system, a first plurality of content collections from the plurality of content messages, each content collection of the first plurality of content collections comprising associated content from one or more content messages of the plurality of content messages;
filtering the content messages based on the quality metrics and predefined quality thresholds;
communicating, from the server computer system to the first mobile device, interface information for each content collection of the first plurality of content collections;
receiving, from the first mobile device, a selection of a first content collection of the first plurality of content collections; receiving, from the first mobile device, an autoforward communication associated with completion of a presentation of the first content collection at the first mobile device;
generating curated content collections from filtered content messages; and
and automatically communicating a second content collection from the server computer system to the first mobile device for automatic display on the first mobile device in response to receipt of the selection and the autoforward communication,
communicating the curated collections to client devices for presentation.
wherein the second content collection is selected automatically by the server computer system, and wherein a respective piece of content is sorted for presentation based on an individual quality score and a weighting based on at least one of content, location, or time associated with the respective piece of content, and
wherein each individual quality score is based on a set of quality metrics applied to the respective piece of content.
It is clear that all of the elements of the instant application 19/078595 (herein ‘595) claim 11 are to be found in U.S. Patent 12,301,650 (herein ‘650) claims 1 (as the instant application ‘595 claim 11 fully encompasses Patent ‘650 claims 1). The difference between ‘595 claim 11 and ‘650 claim 1 lies in the fact that the ‘650 claim includes many more elements and is thus much more specific. Thus, the invention of claims 1 of the 650 patent is in effect a “species” of the “generic” invention of ‘595 claim 11. It has been held that the generic invention is “anticipated” by the “species”. See In re Goodman, 29 USPQ2d 2010 (Fed. Cir. 1993). Since the ‘595 claim 11 is anticipated by claims 1 of ‘650, it is not patently distinct from ‘650 claims 1.
Claim Analysis - 35 USC § 101 (Judicial Exception)
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefore, subject to the conditions and requirements of this title.
The claimed invention is directed to statutory subject matter and no 35 USC 101 rejection is applied for the judicial exception. The claims are directed to non-abstract improvements in computer related technology. The claimed subject matter is integrated into a practical application under Prong 2 of the Step 2A analysis described in MPEP 2016.04(d). A claim is non-statutory when it is directed to a judicial exception (e.g. either one of mathematical concepts, mental processes, or certain methods of organizing human activity) without significantly more. The claimed invention is not directed to a judicial exception. Instead, the claimed invention is directed to a technological improvement for content filtering in a messaging system where embodiments teach analyzing a plurality of content messages to determine quality scores for each content message so that the content messages can be curated into a collection and communicated to client devices for presentation. The claimed invention determines the quality score by performing motion blur estimation on the image content, analyzing audio content for quality metrics, and applying neural network-based content filtering. The ordered steps of the claim language impose meaningful limits on the scope of the claims and provides an improvement for curating short content items by optimizing the quality of the content being shown. Therein the claimed invention is statutory under 35 USC 101
Claim Rejections - 35 USC § 102
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 following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claim(s) 11, 13 – 15, 17 and 20 are rejected under 35 U.S.C. 102(a)(2) as being anticipated by Collins III et al. (U.S. 2014/0156677 A1; herein referred to as Collins)
In regard to claim 11, Collins teaches A method for automated content curation (see abstract “ . . . A presentation order of curated media items that are represented by meta data in a playpack is determined by evaluating a platform score, a viewer score, a content popularity score, a content quality score, and a scatter algorithm. . . .”) , comprising:
receiving content messages from a plurality of client devices (see ¶¶ [0015-0017] “ . . . Methods and systems of ordering curated content for use in a Playpack may include a method of determining a measure of media items that are accessible via metadata stored in a Playpack, including determining a platform assessment score that includes a screen score, a network factor, a network speed, and a touch screen score; determining a viewer assessment score that includes a presentation score, a duration score, a taxonomy score, and repeated view tolerance score; determining a content assessment score that includes a popularity weighting score and a quality score; and ordering the media items based on content assessment score and viewer assessment score. In this method, the quality score is based on comparing platform assessment score factors with media item meta data. Methods and systems of ordering curated content for use in a Playpack mode may include a method of determining a presentation order of media items that are accessible via metadata stored in a Playpack that may include determining a platform assessment score that includes a screen score, a network factor, a network speed, and a touch screen score; determining a viewer assessment score that includes a presentation score, a duration score, a taxonomy score, and repeated view tolerance score; determining a content popularity weighting score; determining a content quality score that is based on comparing platform assessment score factors with media item meta data; ordering the media items based on content popularity weighting score, content quality score and viewer assessment score; and applying a scatter algorithm to re-order the ordered media items based on a viewer repetition tolerance factor. Methods and systems of ordering curated content for use in a Playpack mode may include a media item recommendation engine for recommending a next item to present to a child that is selected from a plurality of media items that are represented by metadata stored in a Playpack that may include a platform assessment score that includes a screen score, a network factor, a network speed, and a touch screen score; a viewer assessment score that includes a presentation score, a duration score, a taxonomy score, and repeated view tolerance score; a content popularity weighting score; a content quality score that is based on comparing platform assessment score factors with media item meta data; a media item presentation order sorting facility for ordering the media items based on content popularity weighting score, content quality score and viewer assessment score; and wherein a media item that appears at the top of the presentation order is the recommended next item to present to the child. . . .”) . . .”);
analyzing the content messages using computer vision to determine image quality metrics (see ¶ [0042] “ . . . A presentation order of curated media items that are represented by meta data in a playpack may be determined by evaluating a platform score, a viewer score, a content popularity score, a content quality score, and a scatter algorithm. Interacting with ordered curated content and use of a Playpack capability may be monitored with the safe child behavior monitoring and data usage capabilities described herein. . ..”)
filtering the content messages based on the quality metrics and predefined quality thresholds (e.g. over zero) (see ¶ ¶ [0088-0090] “. . . Any Media Items with a "Quality Score" of zero are treated as non-existent for purposes of establishing desirability. If a Media Item is Interactive, as opposed to video or audio, then if the Display Device does not have a touch-screen, then that Media Item is considered non-existent for purposes of establishing desirability. All Media Items with non-zero Quality Scores are ordered, highest score to lowest, where the highest are the most desirable, and the lowest are least desirable for this Viewer at this time. . . .” see ¶ [0083] “. . . The remaining steps in the method require iteration over each Media Item in the Playpack. The goal of repeated iterations is to order the Media Items in the Playpack according to desirability. The first iteration (M) applies a centrally-determined "Popularity Weighting" ("PW"). PW is an integer from 0 to 100, which is applied first at Playpack creation, and then again after each addition of one or more Media Items to a Playpack. It gives the author of the Playpack a means to order the Media Items within the Playpack. . ..”))
generating curated content collections (e.g. playback) from filtered content messages (see ¶¶ [0045-0046] “. . . his disclosure describes and depicts a system for organizing and managing one or more digital Media Items. FIG. 2 shows the relationships between the components of this system. To create a Playpack, a user employs one or more Central Authoring and Distribution servers (A). "Authoring" a Playpack consists of selecting one or more Media Items and recording Meta-data about each one inside the Playpack. This is accomplished by using an interactive program on an Authoring Computing Device (B), to enter information via a network connection to a Distribution Server (C) . . . “ see ¶ [0083] “ . . . Media Items with a lower weight number are less desirable overall, and ones with higher are more desirable. Media Items with identical weightings are deemed equally desirable overall. For example, in the previous example of a Playpack with Media Items about volcanoes, a recent volcano eruption might result in new Media Items becoming available, which are of immediate interest to all viewers because of the current nature of the Media Item. Those Media Items would receive a higher PW by the Playpack author when the Playpack is updated. . ..”); and
communicating the curated collections to client devices for presentation (see ¶ ¶ [0092-0093] “ . . .. That score is now used to order the list of Media Items according to the "closeness" of fit. A higher Taxonomy Score indicates a "closer fit" relative to the other Media Items in the Playpack. That sorted list is then re-ordered according to one or more "scatter" algorithms to reduce repetition, according to Viewer tolerance for repetition as discussed above. The result of this method is an ordered list of Media Items in this Playpack, representative of the likelihood of each Media Item being desirable to this Viewer at this time. The Display Device then presents these Media Items in order of desirability to the Viewer. . . “)
In regard to claim 13, Collins teaches wherein filtering content messages comprises:
comparing quality metrics to threshold values (see Collins ¶¶ [0081-0082] “. . . Once we have assembled the above metrics for the first two phases of this method, we can apply those metrics to the content in this Playpack to achieve a ranked list of Media Items, ordered by desirability for this Viewer at this time (K). The first step of this phase is to apply any central updates (L) to the Playpack. Each Playpack contains one or more Media Items. Those Media Items are augmented with new Media Items from time to time. For example, a particular Playpack might initially be transmitted to all Display Devices with ten Media Items in it. Each week for four weeks, an additional two Media Items may be added to the Playpack. At the end of four weeks, the Playpack contains eighteen Media Items. . . “)
selecting content exceeding quality thresholds (see ¶ [0090] “. . . All Media Items with non-zero Quality Scores are ordered, highest score to lowest, where the highest are the most desirable, and the lowest are least desirable for this Viewer at this time . . .”); and
organizing selected content into collections (see ¶ [0092] “. . . That score is now used to order the list of Media Items according to the "closeness" of fit. A higher Taxonomy Score indicates a "closer fit" relative to the other Media Items in the Playpack. That sorted list is then re-ordered according to one or more "scatter" algorithms to reduce repetition . . . “)
In regard to claim 14, Collins teaches further comprising:
analyzing audio components of video content (see Collins ¶ [0043] “. . . Referring to FIG. 1, a functional diagram of Playpack components, a Playpack organizes digital media for viewing and interaction by a user. A Playpack organizes at least two types of digital media, "Online" media (G), and "Local" media (H). "Media" refers to digital representations of images, audio, interactive graphic programs, video, games, books, and the like. A Media Item is considered to be "Online" when it is available by using the display system to access a remote server through a network connection. A Media Item is considered to be "Local" when it is available by accessing storage media with the display system without requiring a use of the network connection. A Playpack uses a Media Cache Manager (F) to provide a consistent means of accessing both Online (G) and Local (H) Media Items. In this way, all Media Items are accessed in a substantially similar manner, while the Playpack determines which Media Items can be accessed and displayed. . ..”) ;
generating audio quality scores (see Collins ¶ [0042] “. . . A presentation order of curated media items that are represented by meta data in a playpack may be determined by evaluating a platform score, a viewer score, a content popularity score, a content quality score, and a scatter algorithm. Interacting with ordered curated content and use of a Playpack capability may be monitored with the safe child behavior monitoring and data usage capabilities described herein.; and
combining audio and visual quality scores (see Collins ¶ [0052] “. . . the methods for determining the most desirable media item(s) may be constructed as a media item recommendation engine. Such a recommendation engine may involve developing a uniform description of attributes for media items (e.g. in a library). In particular the uniform description methodology would facilitate describing the items according to how a child might see the item. Such a uniform description approach may be called "a taxonomy," because it aims to develop a formal language uniquely describing the elements of meaning for media items to be accessed through a Playpack. . . .”)
In regard to claim 15, Collins teaches A non-transitory computer-readable medium storing instructions that, when executed, cause a computer system to (see ¶ [0150] “. . . The computer software, program codes, and/or instructions may be stored and/or accessed on machine readable media that may include: computer components, devices, and recording media that retain digital data used for computing for some interval of time; semiconductor storage known as random access memory (RAM); mass storage typically for more permanent storage, such as optical discs, forms of magnetic storage like hard disks, tapes, drums, cards and other types; processor registers, cache memory, volatile memory, non-volatile memory; optical storage such as CD, DVD; removable media such as flash memory (e.g. USB sticks or keys), floppy disks, magnetic tape, paper tape, punch cards, standalone RAM disks, Zip drives, removable mass storage, off-line, and the like; other computer memory such as dynamic memory, static memory, read/write storage, mutable storage, read only, random access, sequential access, location addressable, file addressable, content addressable, network attached storage, storage area network, bar codes, magnetic ink, and the like . . . “):
analyze received content messages using quality detection models (see ¶¶ [0015-0017], ¶ [0042] as described for the rejection of claim 1 and is incorporated herein);
filter analyzed messages based on quality scores and contextual factors (see ¶ ¶ [0088-0090] as described for the rejection of claim 11 and is incorporated herein); ;
generate curated content collections from filtered messages (see ¶¶ [0045-0046] ¶ [0083] as described for the rejection of claim 11 and is incorporated herein); and
communicate the collections for presentation on client devices (see ¶ ¶ [0092-0093] as described for the rejection of claim 11 and is incorporated herein)
In regard to claim 17, Collins teaches wherein generating curated content collections comprises:
selecting content messages exceeding quality thresholds (see ¶ [0090] “. . . All Media Items with non-zero Quality Scores are ordered, highest score to lowest, where the highest are the most desirable, and the lowest are least desirable for this Viewer at this time . . .”) ;
weighting selected messages based on content type, location, and time (see ¶ [0083] “ . . . The remaining steps in the method require iteration over each Media Item in the Playpack. The goal of repeated iterations is to order the Media Items in the Playpack according to desirability. The first iteration (M) applies a centrally-determined "Popularity Weighting" ("PW"). PW is an integer from 0 to 100, which is applied first at Playpack creation, and then again after each addition of one or more Media Items to a Playpack. It gives the author of the Playpack a means to order the Media Items within the Playpack. . . .”); and
organizing weighted messages into collections (see ¶ [0092] “. . . That score is now used to order the list of Media Items according to the "closeness" of fit. A higher Taxonomy Score indicates a "closer fit" relative to the other Media Items in the Playpack. That sorted list is then re-ordered according to one or more "scatter" algorithms to reduce repetition . . . “).
In regard to claim 20, Collins teaches wherein filtering analyzed messages comprises:
comparing quality scores to threshold values (see Collins ¶¶ [0081-0082] “. . . Once we have assembled the above metrics for the first two phases of this method, we can apply those metrics to the content in this Playpack to achieve a ranked list of Media Items, ordered by desirability for this Viewer at this time (K). The first step of this phase is to apply any central updates (L) to the Playpack. Each Playpack contains one or more Media Items. Those Media Items are augmented with new Media Items from time to time. For example, a particular Playpack might initially be transmitted to all Display Devices with ten Media Items in it. Each week for four weeks, an additional two Media Items may be added to the Playpack. At the end of four weeks, the Playpack contains eighteen Media Items. . . “) ;
selecting messages exceeding thresholds (see ¶ [0090] “. . . All Media Items with non-zero Quality Scores are ordered, highest score to lowest, where the highest are the most desirable, and the lowest are least desirable for this Viewer at this time . . .”); and
organizing selected messages based on quality scores and contextual weights (see ¶ [0092] “. . . That score is now used to order the list of Media Items according to the "closeness" of fit. A higher Taxonomy Score indicates a "closer fit" relative to the other Media Items in the Playpack. That sorted list is then re-ordered according to one or more "scatter" algorithms to reduce repetition . . . “).
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 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 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 of this title, 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.
Claims 1 - 10, 12, 16, and 18 - 19 are rejected under 35 U.S.C. 103 as being unpatentable over Collins III et al. (U.S. 2014/0156677 A1; herein referred to as Collins) as applied to claims 11, 13 – 15, 17, and 20 in view of Lin (U.S. 2014/0023291 A1; herein referred to as Lin) in further view of De Vries et al. (U.S. 2016/0008526 A1; herein referred to as De Vries) in further view of Pau et al. (US 2015/0154456 A1; herein referred to as Pau)
In regard to claim 1, Collins teaches A method for content filtering in a messaging system (see abstract “ . . .a presentation order of curated media items that are represented by meta data in a playpack is determined by evaluating a platform score, a viewer score, a content popularity score, a content quality score, and a scatter algorithm . . .”) , comprising: analyzing, by a server computer system (see Fig. 11, Fig. 12 ¶¶ [0096-0097] “ . . . FIG. 11 shows a typical configuration of Remote Mobile Devices (1 & 2) connected through a network to a Central Server (3). In this configuration, users install and operate application software on the Remote Mobile Device. FIG. 12 shows the typical configuration for a single Remote Mobile Device (4) connected through the network to a Central Server (5). In this configuration, the user has installed three applications (1,2,3). These applications operate on the device performing various functions for the user, such as taking and displaying photos, searching the Internet for information, or connecting to commerce sites to buy and sell merchandise. As a by-product of using these applications, the applications can record and store information about how the application was used, and what interactions uses have with the application. This information can contain time-stamps, to show the exact sequence in which the user interactions transpired. This data, data about the user interaction with applications on the device, is transmitted to a Central Server (5), which stores it in a central database (6). . . .”), a plurality of content messages to determine quality scores for each content message (see ¶¶ [0015-0017] “ . . . Methods and systems of ordering curated content for use in a Playpack may include a method of determining a measure of media items that are accessible via metadata stored in a Playpack, including determining a platform assessment score that includes a screen score, a network factor, a network speed, and a touch screen score; determining a viewer assessment score that includes a presentation score, a duration score, a taxonomy score, and repeated view tolerance score; determining a content assessment score that includes a popularity weighting score and a quality score; and ordering the media items based on content assessment score and viewer assessment score. In this method, the quality score is based on comparing platform assessment score factors with media item meta data. Methods and systems of ordering curated content for use in a Playpack mode may include a method of determining a presentation order of media items that are accessible via metadata stored in a Playpack that may include determining a platform assessment score that includes a screen score, a network factor, a network speed, and a touch screen score; determining a viewer assessment score that includes a presentation score, a duration score, a taxonomy score, and repeated view tolerance score; determining a content popularity weighting score; determining a content quality score that is based on comparing platform assessment score factors with media item meta data; ordering the media items based on content popularity weighting score, content quality score and viewer assessment score; and applying a scatter algorithm to re-order the ordered media items based on a viewer repetition tolerance factor. Methods and systems of ordering curated content for use in a Playpack mode may include a media item recommendation engine for recommending a next item to present to a child that is selected from a plurality of media items that are represented by metadata stored in a Playpack that may include a platform assessment score that includes a screen score, a network factor, a network speed, and a touch screen score; a viewer assessment score that includes a presentation score, a duration score, a taxonomy score, and repeated view tolerance score; a content popularity weighting score; a content quality score that is based on comparing platform assessment score factors with media item meta data; a media item presentation order sorting facility for ordering the media items based on content popularity weighting score, content quality score and viewer assessment score; and wherein a media item that appears at the top of the presentation order is the recommended next item to present to the child. . . .”) . . .”) , wherein determining the quality scores comprises (see ¶ [0042] “ . . . A presentation order of curated media items that are represented by meta data in a playpack may be determined by evaluating a platform score, a viewer score, a content popularity score, a content quality score, and a scatter algorithm. Interacting with ordered curated content and use of a Playpack capability may be monitored with the safe child behavior monitoring and data usage capabilities described herein. . . .”) :
generating filtered content collections (e.g. playback) by (see ¶¶ [0045-0046] “ . . . his disclosure describes and depicts a system for organizing and managing one or more digital Media Items. FIG. 2 shows the relationships between the components of this system. To create a Playpack, a user employs one or more Central Authoring and Distribution servers (A). "Authoring" a Playpack consists of selecting one or more Media Items and recording Meta-data about each one inside the Playpack. This is accomplished by using an interactive program on an Authoring Computing Device (B), to enter information via a network connection to a Distribution Server (C) . . . “)
selecting content messages having quality scores above a threshold (e.g. over zero)(see ¶ ¶ [0088-0090] “ . . . Any Media Items with a "Quality Score" of zero are treated as non-existent for purposes of establishing desirability. If a Media Item is Interactive, as opposed to video or audio, then if the Display Device does not have a touch-screen, then that Media Item is considered non-existent for purposes of establishing desirability. All Media Items with non-zero Quality Scores are ordered, highest score to lowest, where the highest are the most desirable, and the lowest are least desirable for this Viewer at this time. . . .”)
weighing selected content messages based on content type, location, and time (see ¶ [0083] “ . . . The remaining steps in the method require iteration over each Media Item in the Playpack. The goal of repeated iterations is to order the Media Items in the Playpack according to desirability. The first iteration (M) applies a centrally-determined "Popularity Weighting" ("PW"). PW is an integer from 0 to 100, which is applied first at Playpack creation, and then again after each addition of one or more Media Items to a Playpack. It gives the author of the Playpack a means to order the Media Items within the Playpack. . . .”); and
grouping weighted content messages into collections based on the weightings (see ¶ [0083] “ . . . Media Items with a lower weight number are less desirable overall, and ones with higher are more desirable. Media Items with identical weightings are deemed equally desirable overall. For example, in the previous example of a Playpack with Media Items about volcanoes, a recent volcano eruption might result in new Media Items becoming available, which are of immediate interest to all viewers because of the current nature of the Media Item. Those Media Items would receive a higher PW by the Playpack author when the Playpack is updated. . . .”);
communicating the filtered content collections to client devices for presentation (see ¶ ¶ [0092-0093] “ . . .. That score is now used to order the list of Media Items according to the "closeness" of fit. A higher Taxonomy Score indicates a "closer fit" relative to the other Media Items in the Playpack. That sorted list is then re-ordered according to one or more "scatter" algorithms to reduce repetition, according to Viewer tolerance for repetition as discussed above. The result of this method is an ordered list of Media Items in this Playpack, representative of the likelihood of each Media Item being desirable to this Viewer at this time. The Display Device then presents these Media Items in order of desirability to the Viewer. . . “).
Collins fails to explicitly teach
However Lin teaches performing motion blur estimation on image content )see ¶ [0035] “ . . . Some embodiments are designed to handle out-of-focus blur and small motion blur. The out-of-focus blur can be modeled by either a 2D circular-disk kernel or a 2D isotropic Gaussian kernel with zero-mean and 1D standard deviation .phi. (which can be approximated as recursive filtering of the image with an elementary 2D Gaussian kernel with a small .sigma.), and the small motion blur can be modeled by either a 1D uniform kernel or 1D Gaussian kernel with zero-mean and standard deviation .phi. with a known direction (which can be modeled as a recursive filtering of the image with a directional 1D Gaussian kernel with a small .sigma.). . . .”) ;
It would have been obvious to one with ordinary skill in the art before the effective filing date of the applicant’s invention to incorporate a system and method to extract portions of video data for image deblurring The extracted portions of video data are used to test for motion blur within video frames for quality control purpose, as taught by Lin into a system and method for creating a presentation order of curated media items that are represented by meta data in a play pack and is determined by evaluating a platform score, a viewer score, a content popularity score, a content quality score, and a scatter algorithm, as taught by Collins. Such incorporation enables media items to be evaluated based on the quality of the video frames.
The combination of Collins and Lin fails to explicitly teach
However, De Vries analyzing audio content for quality metrics (see ¶ [0043] “ . . . The extent of "user satisfaction" cannot be determined entirely through objective metrics such as signal-to-noise ratio or loudness. Assuming that there exists an `internal` metric in a user's brain that corresponds to his appreciation of the received sound, this "sound quality" metric may be modelled by a user satisfaction or utility function U(y;.omega.), where y represents an audio signal and .omega..di-elect cons..OMEGA. the tunable parameters of the utility model. The term "utility" is from Decision Theory terminology. Since y=F(x;.theta.), U(y;.omega.)=U(x;.theta.,.omega.). The last expression is useful, since it shows the implicit dependency of the utility on the hearing aid algorithm parameters E. In the following U(y.sub.1)>U(y.sub.2) indicates that audio signal y.sub.1 is preferred to y.sub.2. . . .”) ;
It would have been obvious to one with ordinary skill in the art before the effective filing date of the applicant’s invention to incorporate a system and method for effective estimation of signal processing parameters in audio, that maybe distinguishable through human hearing, as taught by De Vries, inro a system and method for creating a presentation order of curated media items that are represented by meta data in a play pack and is determined by evaluating a platform score, a viewer score, a content popularity score, a content quality score, and a scatter algorithm, and further extracting portions of video data and used to test for motion blur within video frames for quality control purpose, as taught by the combination of Collins and Lin. Such incorporation provides testing the audio content of the media item for quality control.
The combination of Collins, Lin and De Vries fails to explicitly teach
However, Pau teaches and applying neural network-based content filtering (see ¶ [0073] “ . . . individual classifiers are typically based on minimally complex mathematics like Support Vector Machines or multilayer feed-forward Neural Networks whose classification models are very compact and minimally memory-demanding . . .”);
It would have been obvious to one with ordinary skill in the art before the effective filing date of the applicant’s invention to incorporate a system and method for extraction of descriptors from video content, includes the following steps: a Key Frame Extracting step, applying a local descriptors-based approach to select pictures of the incoming video as key frames that are representative of a temporal region of the video which is visually homogeneous; a Content Analysis step, analyzing the content of the key frames and classifying image patches of the key frames as interesting or not for the extraction of descriptors; a Descriptors Extracting step, extracting compact descriptors from the selected key frames, and defining a set of surrounding images also on the basis of input received from the Content Analysis step, as taught by Pau, inro a system and method for creating a presentation order of curated media items that are represented by meta data in a play pack and is determined by evaluating a platform score, a viewer score, a content popularity score, a content quality score, and a scatter algorithm, and further extracting portions of video data and used to test for motion blur within video frames and audio data tested using signal processing techniques for quality control purpose, as taught by the combination of Collins, Lin, and De Vries. Such incorporation provides neural analysis for determining the quality of the media content being curated.
In regard to claim 2, the combination of Collins, Lin, De Vries, and Pau teaches wherein performing motion blur estimation comprises:
calculating energy gradients on detected edges of image content (see Lin ¶ [0060] “ . . . there is also a prior term added to the data fidelity term, which encourages sparsity of gradient energy over the entire image or minimizes total variation, embodiments encourage the sparsity implicitly by bypassing the addition of higher frequency components on low-contrast, smooth regions, which also significantly accelerates the overall patch search process. . . .”); and
identifying video frames with motion blur above a threshold amount (see Lin ¶ [0060] “ . . . The low contrast and smooth regions are identified by thresholding either variance of the patches or the distance to the nearest neighbor patches. Some embodiments also explicitly compute the full cost function (sum of data term and sparsity term) in each iteration to sparsify the .DELTA. and then perform frequency band transfer in Algorithm 1 and 2, . . . “).
The motivation to combine the references is described for the rejection of claim 1 and is incorporated herein. Additionally, Lin employes techniques for identifying regions of blur in the image.
In regard to claim 3, the combination of Collins, Lin, De Vries, and Pau teaches wherein applying neural network-based content filtering (see ¶ [0013] “ . . . Upon successfully identifying the matching images, the best image is marked as the current key-node, and the set of images in-play is reduced to only those images that are connected by a path in the database. The database of images is organized as follows: V is a collection of images; G is an undirected graph where images forms the nodes in the graph, and the edges describe the relationships between the images. An edge between two images indicates a geometric relationship when these two images can be related through standard pairwise image matching. Each image is also further identified with one or more identifiers and two images sharing the same identifier are also connected by an additional edge. This organization is similar to a graph of images constructed for hierarchical browsing purposes. . . .”) comprises:
extracting features from content using a feed-forward artificial neural network (see Pau ¶¶ [0033-0036]” . . . a method for extraction of semantic descriptors from video content comprises the following main steps: [0034] a Key Frame Extracting step, applying a local descriptors-based approach to select pictures of the incoming video as key frames that are representative of a temporal region of the video which is visually homogeneous; [0035] a Content Analysis step, analyzing the content of said key frames and classifying image patches of said key frames as semantically interesting or not for said extraction of descriptors; [0036] a Descriptors Extracting step, extracting compact descriptors from said selected key frames, and defining a set of surrounding images also on the basis of input received from said Content Analysis step; . . . “) ;
identifying desirable elements of images based on a learning set (see Pau ¶¶ [0020-0022] “ . . . a Key Frame Extractor block, which uses a local descriptors-based approach to select pictures of the incoming video as key frames that are representative of a temporal region of the video which is visually homogeneous; [0021] a Content Analyzer block, which analyses the content of said key frames and classifies image patches of said key frames as semantically interesting or not for said extraction of compact descriptors; [0022] a Descriptors Extractor block, which extracts said compact descriptors from said selected key frames, and defines a set of surrounding images also on the basis of input received from said Content Analyzer block ; and
assigning an interestingness score based on a neural network analysis (see Pau ¶ [0082] “ . . . The role of the Ensemble Classifier block is then to provide as output a probability estimation about the relevance of each of the input patches. This is done by a weighted linear combination of the classification decisions using the classification confidence scores as weights. . . .”0
The motivation to combine the references is described for the rejection of claim 1 and is incorporated herein. Additionally, Pau determines desirability of the image using neural techniques.
In regard to claim 4, the combination of Collins, Lin, De Vries, and Pau teaches wherein weighting selected content messages comprises: applying different weights to different images based on content type (see Collins ¶ [0015] “ . . .determining a viewer assessment score that includes a presentation score, a duration score, a taxonomy score, and repeated view tolerance score; determining a content assessment score that includes a popularity weighting score and a quality score; and ordering the media items based on content assessment score and viewer assessment score. In this method, the quality score is based on comparing platform assessment score factors with media item meta data. ;
In regard to claim 5, the combination of Collins, Lin, De Vries, and Pau teaches wherein analyzing audio content comprises: evaluating dynamic range (see De Vries ¶ [0045] “ . . . the utility function U(y,.omega.) is different for each user (and may even change over time for a single user). All measurable user data relevant to a utility function are collected in a parameter vector .alpha..di-elect cons.A. The vector .alpha., in the following denoted the auditory profile, portrait or signature, includes data such as the audiogram, SNR-loss, dynamic range, lifestyle parameters and possibly measurements about a user's cochlear, binaural or central hearing deficit . . .”) ;
analyzing noise levels (see De Vries ¶ [0079] “ . . . the dispenser measures relevant user information (such as the audiogram and/or a speech-in-noise test) . . “) ; and
determining language (e.g. speech) clarity. (see De Vries ¶ [0044] “ . . . An example for the utility function would be the PESQ function (PESQ=Perceptual Evaluation of Speech Quality), which is an International Telecommunication Union (ITU) standard (ITU-T Recommendation P.862) that assigns a speech quality rating (a value between 1 and 5) to a speech signal. This rating is supposed to correspond to how humans rate the quality of speech signals. The parameters in the PESQ function have been selected so that the output of the PESQ function matches the average human responses as closely as possible. According to the present invention, the parameters of the PESQ function are allowed to vary, and the uncertainties relating to values of the utility parameters w is expressed by a probability distribution function (PDF) P(.omega.|.alpha.). Over time, information about the parameters .omega. of the utility function is gained through experiments (D) and hereby information is also gained about the (personal) utility function U(y;.omega.). Other utility functions may be PAQM, PSQM, NMR, PERCEVAL, DIX, OASE, POM, PEAQ, etc. Another alternative is the speech intelligibility metric disclosed in: "Coherence and the speech intelligibility index", by James M. Kates et. al. in J. Acoust. Soc. Am. 117 (4), 1 April 2005. . . .”).
The motivation to combine the references is described in the rejection of claim 1 and is incorporated herein. Additionally, De Vries provides measurements for audio media content.
In regard to claim 6, Collins teaches A system for content filtering in a messaging platform (see abstract as described for the rejection of claim 1 and is incorporated herein) , comprising: one or more processors (see ¶ [0141] “ . . . The methods and systems described herein may be deployed in part or in whole through a machine that executes computer software, program codes, and/or instructions on a processor. The processor may be part of a server, client, network infrastructure, mobile computing platform, stationary computing platform, or other computing platform. A processor may be any kind of computational or processing device capable of executing program instructions, codes, binary instructions and the like. . . .”) ; and
memory storing instructions that, when executed by the one or more processors ((see ¶ [0143] “ . . . The processor may access a storage medium through an interface that may store methods, codes, and instructions as described herein and elsewhere. The storage medium associated with the processor for storing methods, programs, codes, program instructions or other type of instructions capable of being executed by the computing or processing device may include but may not be limited to one or more of a CD-ROM, DVD, memory, hard disk, flash drive, RAM, ROM, cache and the like. . . .”), cause the system to:
analyze received content messages using quality metrics (see ¶¶ [0015-0017], ¶ [0042] as described for the rejection of claim 1 and is incorporated herein)
generate filtered content collections (e.g. playback) by (see ¶¶ [0045-0046] as described for the rejection of claim 1 and is incorporated herein) selecting and weighting content based on quality scores and contextual factors (see ¶ [0083], ¶ ¶ [0088-0090] as described for the rejection of claim 1 and is incorporated herein) ; and
communicate the filtered collections for presentation on client devices (see ¶ ¶ [0092-0093] as described for the rejection of claim 1 and is incorporated herein).
Collins fails to explicitly teach
including motion blur detection, audio quality analysis, and neural network-based content filtering;
However, Lin teaches including motion blur detection (see ¶ [0035] as described for the rejection of claim 1 and is incorporated herein)
The motivation to combine Lin with Collins is described for the rejection of claim 1 and is incorporated herein.
The combination of Collins and Lin fails to explicitly teach audio quality analysis, and neural network-based content filtering;
However, De Vries teaches audio quality analysis (see ¶ [0043] as described for the rejection of claim 1 and is incorporated herein)
The motivation to combine De Vries with the combination of Collins and Lin is described for the rejection of claim 1 and is incorporated herein.
The combination of Collins, Lin, and De Vries fails to explicitly teach
However, Pau teaches and neural network-based content filtering (see ¶ [0073] as described for the rejection of claim 1 and is incorporated herein)
The motivation to combine Pau with the combination of Collins, Lin, and De Vries is described for the rejection of claim 1 and is incorporated herein.
In regard to claim 7, the combination of Collins, Lin, De Vries, and Pau teaches wherein generating filtered content collections comprises:
Identify video frames with camera motion or shake above a threshold (see Lin ¶ [0002] “ . . . This profusion of low-quality photographs has made improvement in basic image restoration tasks, such as image deblurring, sharpening, and denoising, substantially potentially valuable for users of graphics software. More specifically, image deblurring and sharpening is a long-standing, fundamental problem in digital image restoration, referred to as "a field of engineering that studies methods used to recover an original scene from degraded observations". Image blurs can be divided into two categories: defocus blur and motion blur, and motion blur can be caused by either/both camera shake and/or independent moving objects, e.g. people. . . .”) ;
modify overall quality scores based on identified motion (see Lin ¶ [0035] “ . . . the small motion blur can be modeled by either a 1D uniform kernel or 1D Gaussian kernel with zero-mean and
standard deviation .phi. with a known direction (which can be modeled as a recursive filtering of the image with a directional 1D Gaussian kernel with a small .sigma.). The number of iterations (filtering) N for the recursive filtering can be estimated as N=.left brkt-bot..phi..sup.2/.sigma..sup.2.right brkt-bot., where .left brkt-bot.x.right brkt-bot. computes the largest integer smaller than or equal to x. . . .”) ;
filter out videos with camera motion exceeding the threshold (see Lin ¶ [0040] “ . . . When there is significant noise in the input image, some embodiments apply a modification to the above algorithm by finding k nearest neighbors instead of the single nearest neighbor, and update the high frequency component by a weighted sum of the high frequency components from multiple example patches. Algorithm 2 describes an embodiment of an algorithm for simultaneous denoising and deblurring using k-nearest neighbor patches finding (the pixel intensity and any color channel value is normalized to a range of [0 1]). . . .”).
The motivation to combine Lin with the other references is described for the rejection of claim 1 and is incorporated herein. Additionally, Lin offers solutions to improving quality of poor images.
In regard to claim 8, the combination of Collins, Lin, De Vries, and Pau teaches wherein generating filtered content collections comprises:
analyzing content for compression artifacts (see Lin ¶ [0003] “ . . . Image deblurring is a long-studied problem in image processing and computer vision, but adequate solutions have evaded researchers and developers due to complexity of blur patterns in the image, image noise and non-uniformity of blur kernels in images. Most previously-existing approaches to non-blind deconvolution are either too slow, too sensitive to noise, or prone to ringing artifacts near edges due to loss of information by the underlying blurring process. Often, previously-existing approaches generate artifacts resulting from deconvolution, and/or remove the amplification effects of noise with a post-processing method. More importantly, when blur kernels are not spatially piece-wise constant, the small spatial difference of kernels cause significant artifacts in performing deconvolution . . .”) ;
identifying image quality issues based on the compression artifacts (see Lin ¶ [0034] “ . . . Some embodiments execute a local patch self similarity-based approach for image sharpening and deblurring, which recursively sharpens an input blurry image to obtain a sharp, deblurred result. Some embodiments are orders of magnitude faster than previously-existing blind/non-blind deconvolution algorithms. Experimental results show that some embodiments generate more natural, sharper, artifact-free results than the best available non-blind image deconvolution approach, and are significantly more robust to input image noise and spatial non-uniformity of the blur kernels. While some embodiments are directly applied to image super-resolution by deblurring the bicubic-interpolated upscaled images, embodiments are not so limited. In some embodiments, denoising and deblurring are handled simultaneously. . . . “) ; and
modifying quality scores based on identified artifacts (see Lin ¶ [0043] “ . . . Some embodiments apply the above algorithm for clean input images. When there is significant noise in the input image, some embodiments apply a modification to the above algorithm by finding k nearest neighbors instead of the single nearest neighbor, and update the high frequency component by a weighted sum of the high frequency components from multiple example patches. Algorithm 2 describes an embodiment of an algorithm for simultaneous denoising and deblurring using k-nearest neighbor patches finding (the pixel intensity and any color channel value is normalized to a range of [0 1]). Note that, for a gray-scale image, some embodiments directly run the following algorithm; while for a color RGB image, for efficiency, some embodiments convert the input image to YCbCr space, and run the deblurring algorithm on the luminance channel Y only. Alternatively, some embodiments independently deblur all three color channels, but the computation is 3-times slower and the resulting quality is similar to the former option. . . “)
The motivation to combine Lin with the other references is described for the rejection of claim 1 and is incorporated herein. Additionally, Lin provides techniques for artifact analysis in the sampled image,
In regard to claim 9, the combination of Collins, Lin, De Vries, and Pau teaches wherein analyzing received content messages comprises:
analyzing variance in uniform regions of images for noise artifacts (see Lin ¶ [0034] “ . . Some embodiments execute a local patch self-similarity-based approach for image sharpening and deblurring, which recursively sharpens an input blurry image to obtain a sharp, deblurred result. Some embodiments are orders of magnitude faster than previously-existing blind/non-blind deconvolution algorithms. Experimental results show that some embodiments generate more natural, sharper, artifact-free results than the best available non-blind image deconvolution approach, and are significantly more robust to input image noise and spatial non-uniformity of the blur kernels. While some embodiments are directly applied to image super-resolution by deblurring the bicubic-interpolated upscaled images, embodiments are not so limited. In some embodiments, denoising and deblurring are handled simultaneously. . . .).;
identifying noise associated with camera sensors or optics (see Lin ¶ [0046] “. . . Some embodiments apply the above algorithm for clean input images. When there is significant noise in the input image, some embodiments apply a modification to the above algorithm by finding k nearest neighbors instead of the single nearest neighbor, and update the high frequency component by a weighted sum of the high frequency components from multiple example patches. Algorithm 2 describes an embodiment of an algorithm for simultaneous denoising and deblurring using k-nearest neighbor patches finding (the pixel intensity and any color channel value is normalized to a range of [0 1]). Note that, for a gray-scale image, some embodiments directly run the following algorithm; while for a color RGB image, for efficiency, some embodiments convert the input image to YCbCr space, and run the deblurring algorithm on the luminance channel Y only. Alternatively, some embodiments independently deblur all three color channels, but the computation is 3-times slower and the resulting quality is similar to the former option . . .”); and
adjusting quality scores based on identified noise (e.g. smoothed image) (see Lin ¶ [0065] “. . . high frequency component module 160 performs finding a plurality of near example patches in the smoothed image, extracting a plurality of corresponding patches for the plurality of near examples from a plurality of corresponding locations in the downsized image, computing (frequency band) differences between the plurality of near example patches and the plurality of corresponding patches, computing a denoised center pixel of a source patch from the plurality of example patches in the smoothed image, computing a denoised high frequency component by linearly combining the plurality of the high frequency components from corresponding patch pairs in the smoothed image and the downsized image, and updating a source patch by adding the denoised high frequency component. . . .”).
The motivation to combine Lin with the other references is described for the rejection of claim 1 and is incorporated herein. Additionally, Lin offers techniques to eliminate noise from images in the media content.
In regard to claim 10, the combination of Collins, Lin, De Vries, and Pau wherein the instructions further cause the system to: analyze audio content for dynamic range (see De Vries ¶ [0045] “. . . All measurable user data relevant to a utility function are collected in a parameter vector .alpha..di-elect cons.A. The vector .alpha., in the following denoted the auditory profile, portrait or signature, includes data such as the audiogram, SNR-loss, dynamic range, lifestyle parameters and possibly measurements about a user's cochlear, binaural or central hearing deficit. The audiogram is a recording of the absolute hearing threshold as a function of frequency. SNR loss is the increased dB signal-to-noise ratio required . . .”);
determine audio quality scores based on a dynamic range analysis (see De Vries ¶ [0046] “. . t is relevant to determine a user's satisfaction value for all possible input signals from `the acoustic world` . . .”).; and combine audio quality scores with visual quality scores (see Collins ¶ [0052] “. . . the methods for determining the most desirable media item(s) may be constructed as a media item recommendation engine. Such a recommendation engine may involve developing a uniform description of attributes for media items (e.g. in a library). In particular the uniform description methodology would facilitate describing the items according to how a child might see the item. Such a uniform description approach may be called "a taxonomy," because it aims to develop a formal language uniquely describing the elements of meaning for media items to be accessed through a Playpack. . . .”).
In regard to claim 12, the combination of Collins, Lin, De Vries, and Pau teaches wherein analyzing the content messages comprises:
detecting motion blur (see Lin ¶ [0035] “. . . Some embodiments are designed to handle out-of-focus blur and small motion blur. The out-of-focus blur can be modeled by either a 2D circular-disk kernel or a 2D isotropic Gaussian kernel with zero-mean and 1D standard deviation .phi. (which can be approximated as recursive filtering of the image with an elementary 2D Gaussian kernel with a small .sigma.), and the small motion blur can be modeled by either a 1D uniform kernel or 1D Gaussian kernel with zero-mean and standard deviation .phi. with a known direction (which can be modeled as a recursive filtering of the image with a directional 1D Gaussian kernel with a small .sigma.). . . .”) through energy gradient analysis (see Lin ¶ [0060] “ . . . there is also a prior term added to the data fidelity term, which encourages sparsity of gradient energy over the entire image or minimizes total variation, embodiments encourage the sparsity implicitly by bypassing the addition of higher frequency components on low-contrast, smooth regions, which also significantly accelerates the overall patch search process. . . .”)
analyzing compression artifacts in images and video frames (see Lin ¶ [0034] “. . . Some embodiments execute a local patch self-similarity-based approach for image sharpening and deblurring, which recursively sharpens an input blurry image to obtain a sharp, deblurred result. Some embodiments are orders of magnitude faster than previously-existing blind/non-blind deconvolution algorithms. Experimental results show that some embodiments generate more natural, sharper, artifact-free results than the best available non-blind image deconvolution approach, and are significantly more robust to input image noise and spatial non-uniformity of the blur kernels. While some embodiments are directly applied to image super-resolution by deblurring the bicubic-interpolated upscaled images, embodiments are not so limited. In some embodiments, denoising and deblurring are handled simultaneously. . ..).; and
generating quality scores based on detected issues (see Lin ¶ [0065] “. . . high frequency component module 160 performs finding a plurality of near example patches in the smoothed image, extracting a plurality of corresponding patches for the plurality of near examples from a plurality of corresponding locations in the downsized image, computing (frequency band) differences between the plurality of near example patches and the plurality of corresponding patches, computing a denoised center pixel of a source patch from the plurality of example patches in the smoothed image, computing a denoised high frequency component by linearly combining the plurality of the high frequency components from corresponding patch pairs in the smoothed image and the downsized image, and updating a source patch by adding the denoised high frequency component. . . .”)
The motivation to combine the references is described for the rejection of claim 2 and is incorporated herein.
In regard to claim 16, the combination of Collins, Lin, De Vries, and Pau teaches wherein analyzing received content messages comprises:
detecting motion blur in images and video frames (see Lin ¶ [0035] “. . . Some embodiments are designed to handle out-of-focus blur and small motion blur. The out-of-focus blur can be modeled by either a 2D circular-disk kernel or a 2D isotropic Gaussian kernel with zero-mean and 1D standard deviation .phi. (which can be approximated as recursive filtering of the image with an elementary 2D Gaussian kernel with a small .sigma.), and the small motion blur can be modeled by either a 1D uniform kernel or 1D Gaussian kernel with zero-mean and standard deviation .phi. with a known direction (which can be modeled as a recursive filtering of the image with a directional 1D Gaussian kernel with a small .sigma.). . . .”) ;
analyzing compression artifacts and noise (see Lin ¶ [0034] “. . . Some embodiments execute a local patch self-similarity-based approach for image sharpening and deblurring, which recursively sharpens an input blurry image to obtain a sharp, deblurred result. Some embodiments are orders of magnitude faster than previously-existing blind/non-blind deconvolution algorithms. Experimental results show that some embodiments generate more natural, sharper, artifact-free results than the best available non-blind image deconvolution approach, and are significantly more robust to input image noise and spatial non-uniformity of the blur kernels. While some embodiments are directly applied to image super-resolution by deblurring the bicubic-interpolated upscaled images, embodiments are not so limited. In some embodiments, denoising and deblurring are handled simultaneously. . ..); and
generating quality scores based on detected issues (see Lin ¶ [0065] “. . . high frequency component module 160 performs finding a plurality of near example patches in the smoothed image, extracting a plurality of corresponding patches for the plurality of near examples from a plurality of corresponding locations in the downsized image, computing (frequency band) differences between the plurality of near example patches and the plurality of corresponding patches, computing a denoised center pixel of a source patch from the plurality of example patches in the smoothed image, computing a denoised high frequency component by linearly combining the plurality of the high frequency components from corresponding patch pairs in the smoothed image and the downsized image, and updating a source patch by adding the denoised high frequency component. . . .”).
The motivation to combine the references is described for the rejection of claim 2 and is incorporated herein.
In regard to claim 18, the combination of Collins, Lin, De Vries, and Pau teaches wherein the instructions further cause the computer system to:
analyze audio content for quality metrics including dynamic range and noise levels (see De Vries ¶ [0045], ¶ [0079] as described for the rejection of claim 5 and is incorporated herein);
generate audio quality scores (see ¶ [0043] as described for the rejection of claim 1 and is incorporated herein); and
combine audio quality scores with visual quality scores (see Collins ¶ [0052] as described for the rejection of claim 1 and is incorporated herein).
The motivation to combine the references is described for the rejection of claim 1 and is incorporated herein.
In regard to claim 19, the combination of Collins, Lin, De Vries, and Pau teaches wherein analyzing received content messages comprises:
applying neural network analysis to content (see Pau ¶ [0073] “. . . individual classifiers are typically based on minimally complex mathematics like Support Vector Machines or multilayer feed-forward Neural Networks whose classification models are very compact and minimally memory-demanding . . .”);
identifying desirable content elements (see Pau ¶¶ [0020-0022] “ . . . a Key Frame Extractor block, which uses a local descriptors-based approach to select pictures of the incoming video as key frames that are representative of a temporal region of the video which is visually homogeneous; [0021] a Content Analyzer block, which analyses the content of said key frames and classifies image patches of said key frames as semantically interesting or not for said extraction of compact descriptors; [0022] a Descriptors Extractor block, which extracts said compact descriptors from said selected key frames, and defines a set of surrounding images also on the basis of input received from said Content Analyzer block ;; and
generating quality scores based on identified elements (see Pau ¶ [0082] “. . . The role of the Ensemble Classifier block is then to provide as output a probability estimation about the relevance of each of the input patches. This is done by a weighted linear combination of the classification decisions using the classification confidence scores as weights. . ..”)
The motivation to combine the references is described for the rejection of claim 1 and claim 3 and is incorporated herein.
Conclusion
There are prior art made of record which are not relied upon but are considered pertinent to applicant’s disclosure. They are listed on the PTO-892 accompanying this action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JAMES N FIORILLO whose telephone number is (571)272-9909. The examiner can normally be reached on 7:30 - 5 PM 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, John A. Follansbee can be reached on 571-272-3964. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/JAMES N FIORILLO/Primary Examiner, Art Unit 2444
Prosecution Insights