Lei Zhang

Patent Trials and Appeals BoardAug 2, 2021

2020002095 (P.T.A.B. Aug. 2, 2021)

UNITED STATES PATENT AND TRADEMARK OFFICE UNITED STATES DEPARTMENT OF COMMERCE United States Patent and Trademark Office Address: COMMISSIONER FOR PATENTS P.O. Box 1450 Alexandria, Virginia 22313-1450 www.uspto.gov APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO. 13/545,261 07/10/2012 Lei Zhang 106861-0469 6843 121312 7590 08/02/2021 Foley & Lardner LLP/ Broadcom Corporation 3000 K Street N.W Suite 600 Washington, DC 20007-5109 EXAMINER BENNETT, STUART D ART UNIT PAPER NUMBER 2481 NOTIFICATION DATE DELIVERY MODE 08/02/2021 ELECTRONIC Please find below and/or attached an Office communication concerning this application or proceeding. The time period for reply, if any, is set in the attached communication. Notice of the Office communication was sent electronically on above-indicated "Notification Date" to the following e-mail address(es): cmckenna@foley.com ipdocketing@foley.com PTOL-90A (Rev. 04/07) UNITED STATES PATENT AND TRADEMARK OFFICE BEFORE THE PATENT TRIAL AND APPEAL BOARD Ex parte LEI ZHANG Appeal 2020-002095 Application 13/545,261 Technology Center 2400 Before MAHSHID D. SAADAT, CATHERINE SHIANG, and SCOTT E. BAIN, Administrative Patent Judges. SAADAT, Administrative Patent Judge. DECISION ON APPEAL STATEMENT OF THE CASE Pursuant to 35 U.S.C. § 134(a), Appellant1 appeals from the Examiner’s decision to reject claims 16, 17, 19, 21–23, and 27–37. We have jurisdiction under 35 U.S.C. § 6(b). We REVERSE. 1 We use the word Appellant to refer to “applicant” as defined in 37 C.F.R. § 1.42(a). Appellant identifies the real party in interest as Avago Technologies International Sales Pte. Limited. Appeal Br. 2. Appeal 2020-002095 Application 13/545,261 2 CLAIMED SUBJECT MATTER The claims are directed to “video coding systems and methods that include a single encoding engine that shares video coding information among plural, real-time, parallel encoding operations to provide multiple encoded streams.” Spec. ¶ 10. The disclosure explains that “motion vector search results (e.g., motion vectors, partitions of one coding unit or one macroblock, motion vector search range, etc.)” and “mode decisions, such as inter or intra prediction mode for the coding unit or macroblock, and intra prediction direction if intra prediction is selected for the coding unit or macroblock” are included in the video coding information. Id. Claims 16, 21, and 23, reproduced below, illustrate the claimed subject matter (with disputed limitations emphasized in italics): 16. A method, comprising: receiving, at a single encoding engine, an input video stream; and generating, by the single encoding engine, in parallel, a plurality of streams comprising a first encoded stream according to a first version of profile and level and a second encoded stream according to a second version of the profile and level, the second encoded stream being generated using video coding information used to generate the first encoded stream, and the first version of the profile and level and the second version of the profile and level differing in either the profile, the level, or a combination of both the profile and the level, wherein the generating of the second encoded stream comprises performing motion search on the second encoded stream using a luma bi-prediction size that is limited to a minimum luma bi-prediction size specified for the first encoded stream when the level of the profile and level associated with the second encoded stream is lower than a threshold level and the level of the profile and level associated with the first encoded stream is equal to or above the threshold level. Appeal 2020-002095 Application 13/545,261 3 21. A system, comprising: a single encoding engine configured to: receive an input video stream; and generate, in parallel, a first encoded stream having a first frame rate and a second encoded stream having a second frame rate from the input video stream, the first frame rate being higher than the second frame rate, wherein the generating of the second encoded stream comprises: identifying a first Group of Pictures (GOP) associated with the first encoded stream and identifying a second GOP associated with the second encoded stream; selecting a second picture from the second GOP, the selected second picture having a corresponding first picture in the first GOP; determining that the corresponding first picture in the first GOP uses a first reference picture that is available in the first GOP and absent from the second GOP; and responsive to the corresponding first picture in the first GOP using the first reference picture that is available in the first GOP and absent from the second GOP, deriving motion vectors for the selected second picture in the second GOP by scaling motion vectors associated with the corresponding first picture in the first GOP by: identifying a first temporal distance between the corresponding first picture in the first GOP and a second reference picture which is preceding to the first reference picture that is available in the first GOP and absent from the second GOP, the second reference picture available in both the first GOP and the second GOP, identifying a second temporal distance between the corresponding first picture in the first Appeal 2020-002095 Application 13/545,261 4 GOP and the first reference picture that is available in the first GOP and absent from the second GOP, determining a factor corresponding to a ratio between the first temporal distance and the second temporal distance, and scaling the motion vectors associated with the corresponding first picture in the first GOP to derive the motion vectors for the selected second picture in the second GOP using the determined factor. 23. A method, comprising: receiving, at a single encoding engine, an input video stream; and generating, by the single encoding engine, in parallel, a plurality of streams comprising a first encoded stream having a first version of a video characteristic and a second encoded stream having a second version of the video characteristic, the first version of the video characteristic and the second version of the video characteristic differing in at least a value of the video characteristic, wherein the second encoded stream is generated using video coding information used to generate the first encoded stream, the video coding information comprising a mode decision for inter prediction or intra prediction, and wherein an intra prediction mode decision of each macroblock corresponding to the first encoded stream is used for the second encoded stream when a same intra picture is shared between the first encoded stream and the second encoded stream, the intra prediction mode decision of the respective macroblock comprising a determination of an intra prediction direction for the respective macroblock. Appeal Br. 31–33 (Claims App.). Appeal 2020-002095 Application 13/545,261 5 REFERENCES The prior art relied upon by the Examiner is: Name Reference Date Sarna US 7,809,061 B1 Oct. 5, 2010 Sun US 2006/0013305 A1 Jan. 19, 2006 Cha US 2006/0165301 A1 July 27, 2006 Hannuksela US 2013/0170561 A1 July 4, 2013 Seregin US 2013/0322538 A1 Dec. 5, 2013 H.264 Advanced video coding for generic audiovisual services, Recommendation ITU-T H.264, pp. 1–657 Jan. 2012 REJECTIONS Claims 21 and 22 stand rejected under 35 U.S.C. § 112(a), or pre-AIA 35 U.S.C. § 112, first paragraph, as failing to comply with the written description requirement. Final Act. 8. Claims 16, 17, and 19 stand rejected under 35 U.S.C. § 103(a) as being unpatentable over Sarna and H.264. Final Act. 9–13. Claims 21 and 22 stand rejected under 35 U.S.C. § 103(a) as being unpatentable over Sarna, Sun, and Seregin. Final Act. 13–18. Claims 23, 27–32, and 34–37 stand rejected under 35 U.S.C. § 103(a) as being unpatentable over Sarna and Hannuksela. Final Act. 18–25. Claim 33 stands rejected under 35 U.S.C. § 103(a) as being unpatentable over Sarna, Hannuksela, Sun, and Seregin. Final Act. 25–27. Claims 21 and 22 stand rejected under 35 U.S.C. § 103(a) as being unpatentable over Sarna, Sun, and Cha. Final Act. 27–32. Appeal 2020-002095 Application 13/545,261 6 OPINION We have reviewed the Examiner’s rejections in light of Appellant’s contentions in the Appeal Brief and the Reply Brief that the Examiner has erred, as well as the Examiner’s response to Appellant’s arguments in the Appeal Brief. We concur with the arguments raised by Appellant and emphasize the following. SECTION 112 REJECTION Regarding independent claims 21 and 22, the Examiner finds that the Specification does not disclose the recited “identifying a first temporal distance . . . identifying a second temporal distance . . . and identifying a factor” and therefore, the claims fail to comply with the written description requirement. Final Act. 8. The Examiner specifically finds Concerning claim 21, the specification in ¶¶ [0023] and [0024] define a ratio in terms of distances, but does not disclose the steps of obtaining the information. The amendments to claim 21 incorporate identifying a first temporal distance . . . identifying a second temporal distance . . . and identifying a factor . . . which is not disclosed in ¶¶ [0023] and [0024]. The disclosure does provide an example as to what the factor by which the motion vector is scaled by corresponds do, but not any means by which the information is obtained. Id. Appellant argues that the Specification supports the recited limitations “in a manner understandable to a person of ordinary skill in the art in a way that shows that the inventor actually invented the claimed invention at the time of filing.” Appeal Br. 8. Appellant points to paragraphs 23, 24, 38, and 39, as well Figures 2A, 2B, 4A, and 4B, and argues that these portions of the disclosure provide support for the claim 21 recited features. Id. Appellant Appeal 2020-002095 Application 13/545,261 7 specifically points to the disclosure of temporal distance between different frames and the ratio of those distances for the specific disclosure in those portions of the Specification and argues that “a person of ordinary skill in the art understands that the ‘factor’ is obtained based on the ‘ratio’ of the temporal distance 208 to the temporal distance 206 (paragraph [0023]).” Id. at 10. We agree with Appellant that Appellant’s original disclosure [suggest citing specific paragraphs] conveys with reasonable clarity to ordinarily skilled artisans that Appellant possessed the claimed invention, including a ratio of temporal distances. “[T]he description requirement does not demand any particular form of disclosure, or that the specification recite the claimed invention in haec verba.” Ariad Pharms., Inc. v. Eli Lilly & Co., 598 F.3d 1336, 1352 (Fed. Cir. 2010) (en banc) (internal citations omitted). Rather, “the test for sufficiency is whether the disclosure of the application relied upon reasonably conveys to those skilled in the art that the inventor had possession of the claimed subject matter as of the filing date.” Ariad, 598 F.3d at 1351. Further, as Appellant argues that “such features are well known in the art, and that their omission from the specification for conciseness, written as it is to a person of ordinary skill in the art, does not suggest that the [inventor] lacked possession of the invention,” Reply Br. 4, the instant disclosure satisfy the written description requirement. See Spec. ¶¶ 23, 24, 38, 39. We therefore agree with Appellant that these descriptions would reasonably convey to a person of ordinary skill in the art the necessary disclosure for “‘identifying a first temporal distance’ between a picture and a reference picture, ‘identifying a second temporal distance’ between the Appeal 2020-002095 Application 13/545,261 8 picture and another reference picture, and ‘determining a factor corresponding to a ratio between the first temporal distance and the second temporal distance.’” See Reply Br. 3–4. Thus, we are persuaded by Appellant’s argument that the Specification provides written description support for the disputed limitation and that the Examiner erred by rejecting the pending claims under 35 U.S.C. § 112(a), or pre-AIA 35 U.S.C. § 112, first paragraph. SECTION 103 REJECTION Claim 23 As reproduced supra, claim 23 requires (emphasis added): the video coding information comprising a mode decision for inter prediction or intra prediction, and wherein an intra prediction mode decision of each macroblock corresponding to the first encoded stream is used for the second encoded stream when a same intra picture is shared between the first encoded stream and the second encoded stream, the intra prediction mode decision of the respective macroblock comprising a determination of an intra prediction direction for the respective macroblock. See Appeal Br. 12. In rejecting claim 23, the Examiner finds that the features of the above-quoted limitation are taught or suggested by the combination of Sarna and Hannuksela. Final Act. 18–20 (citing Sarna 3:12–14, 5:57–6:13, 6:15– 16, 6:29–33, Figs. 3–6; Hannuksela ¶¶ 5, 69, 82, 87; Ans. 14–16. Appellant contends that the Examiner erred in rejecting claim 23 under 35 U.S.C. § 103 because: Appeal 2020-002095 Application 13/545,261 9 As shown in FIG. I of Hannuksela above, a sequence of coded pictures (i.e., I pictures, P pictures, B pictures) is organized into two Group of Pictures (GOPs) 214 arranged in display order which is indicated by the picture order count (POC) 210. See Hannuksela, paragraph [0082]. . . . Hannuksela describes that “Pictures of a higher temporal level may only use pictures of the same or lower temporal level for inter-picture prediction” See Hannuksela, paragraph [0082]. However, neither FIG. 1 of Hannuksela nor other parts of Hannuksela disclose or suggest the above recited features of Claim 23, for at least the reasons that (1) FIG. 1 of Hannuksela, which is cited by the Final Action for disclosing Claim 23, merely describes inter-picture prediction; (2) Hannuksela, as a whole, is completely silent with respect to “an intra prediction mode decision of each macro block” as recited in Claim 23; (3) Hannuksela, as a whole, is completely silent with respect to “a determination of an intra prediction direction for the respective macroblock” as recited in Claim 23; and (4) Hannuksela, as a whole, is completely silent with respect to “an intra prediction mode decision [which comprises a determination of an intra prediction direction] of each macroblock corresponding to the first encoded stream is used for the second encoded stream when a same intra picture is shared between the first encoded stream and the second encoded stream” as recited in Claim 23. Appeal Br. 13–14; see also Reply Br. 5–7. With respect to the disclosure in paragraph 82 of Hannuksela, Appellant argues: The cited lines 15-18 of paragraph [0082] of Hannuksela merely disclose that “[w]ith such a hierarchical coding structure [i.e., TL 0–3], different temporal scalability corresponding to different frame rates can be achieved by discarding pictures of a certain temporal level value and beyond.” This passage at best discloses that a same key picture (i.e., the I/P picture) is shared between pictures of different temporal levels for inter- picture prediction. Appeal Br. 15. Appeal 2020-002095 Application 13/545,261 10 We agree with Appellant that Hannuksela’s disclosure of prediction methods using a reference picture or frame relates to inter prediction and does not teach or suggest the claimed intra prediction mode decision. See Appeal Br. 13–14. The Examiner’s assertion that “[t]he I in ‘I/ P’ means it’s an I-frame which is literally defined as intra-coded picture” and “[i]ntra- prediction, as per the standards which Hannuksela builds upon, is performed according to modes (i.e., DC mode, vertical mode, horizontal mode, etc.) and performed at the macroblock level” (see Ans. 34) are not supported by evidence of record. Accordingly, we agree with Appellant that the Examiner does not provide prima facie support for the rejection. “[T]he examiner bears the initial burden, on review of the prior art or on any other ground, of presenting a prima facie case of unpatentability.” In re Oetiker, 977 F.2d 1443, 1445 (Fed. Cir. 1992). Therefore, on the record before us, we are constrained to find that the Examiner errs in rejecting independent claim 23, as well as claims 27–37 dependent therefrom, as obvious. Claim 16 As reproduced supra, claim 16 requires (emphasis added): generating, by the single encoding engine, in parallel, a plurality of streams comprising a first encoded stream according to a first version of profile and level and a second encoded stream according to a second version of the profile and level, the second encoded stream being generated using video coding information used to generate the first encoded stream, and the first version of the profile and level and the second version of the profile and level differing in either the profile, the level, or a combination of both the profile and the level, Appeal 2020-002095 Application 13/545,261 11 wherein the generating of the second encoded stream comprises performing motion search on the second encoded, stream using a luma bi-prediction size that is limited to a minimum luma bi-prediction size specified for the first encoded stream when the level of the profile and level associated with the second encoded stream is lower than a threshold level and the level of the profile and level associated with the first encoded stream is equal to or above the threshold level. See Appeal Br. 22–23. In rejecting claim 16, the Examiner finds that the features of the above-quoted limitations are taught or suggested by the combination of Sarna and H.264. Final Act. 9–12 (citing Sarna 1:27–28, 3:63–5:4, Figs. 3– 6; H.264 pp. 283, 289, 297, Fig. 7, Tables A-1, A-4); Ans. 5–7. Appellant contends that the Examiner erred in rejecting claim 16 under 35 U.S.C. § 103 because: The H.264 standard merely discloses that the value of the MimLumaBiPredSize is 8X8 for levels 3.1 and above, and is undefined for levels 3 and below. See Table A-4 of the H.264 Standard. However, knowing that the value of the MimLumaBiPredSize is 8X8 for levels 3.1 and above and is undefined for levels 3 and below does not disclose or suggest that the motion search on the second encoded stream is limited to the minimum luma bi-prediction size that is only required for the first encoded stream when the level of the second encoded stream is lower than a threshold level and the level of the first encoded stream is equal to or above the threshold level. Appeal Br. 23; see also Reply Br. 8–9. We agree with Appellant that [O]ne of ordinary skill in the art would not have found it obvious to combine the Sarna (which is silent with respect to even the profile and level) with the H.264 standard (which merely discloses that the value of the MimLumaBiPredSize is Appeal 2020-002095 Application 13/545,261 12 8X8 for levels 3.1 and above, and is undefined for levels 3 and below) to arrive at Claim 16. Appeal Br. 23. The Examiner’s assertion that “Sarna is clear that the H.26# (e.g., H.262, H.263, H.264, etc.) formats are supported as H.264’s predecessor, H.263, is expressly disclosed as a supported format (col. 5, l. 66)” (see Ans. 48) is not supported by evidence of record. Accordingly, we agree with Appellant that the Examiner does not provide prima facie support for the rejection. See In re Oetiker, 977 F.2d at 1445. Therefore, we are persuaded that the Examiner errs in rejecting independent claim 16, as well as claims 17 and 19 dependent therefrom, as obvious. Claim 21 Seregin In rejecting claim 21, the Examiner finds that Sarna discloses the recited encoding engine that receives a video input and generates the first and second encoded streams. Final Act. 13–14 (citing Sarna 1:27, 6:1–2, 15–16, 29–33, Figs. 3–6). The Examiner further finds Sun discloses the steps of “identifying a first Group of Pictures (GOP)” and “determining that the corresponding first picture in the first GOP uses a first reference picture that is available in the first GOP and absent from the second GOP” included in generating the second encoded stream. Final Act. 15–16 (citing Sun ¶ 21, Fig. 1) (emphasis omitted). For the remaining limitations of claim 21, the Examiner relies on the disclosure of the motion vector (MV) scaling in Seregin and finds that “it would have been obvious to a person having ordinary skill in the art to have scaled a motion vector utilizing the temporal Appeal 2020-002095 Application 13/545,261 13 distance, as taught by Seregin ([0039]), in Sarna and Sun’s joint invention.” Final Act. 17–18 (citing Seregin ¶¶ 5, 39). Appellant contends that the Examiner erred in rejecting claim 21 under 35 U.S.C. § 103 because Seregin fails to teach or suggest the following limitations: identifying a first temporal distance between the corresponding first picture in the first GOP and a second reference picture which is preceding to the first reference picture that is available in the first GOP and absent from the second GOP, the second reference picture available in both the first GOP and the second GOP, identifying a second temporal distance between the corresponding first picture in the first GOP and the first reference picture that is available in the first GOP and absent from the second GOP, determining a factor corresponding to a ratio between the first temporal distance and the second temporal distance, and scaling the motion vectors associated with the corresponding first picture in the first GOP to derive the motion vectors for the selected second picture in the second GOP using the determined factor. See Appeal Br. 24. According to Appellant, Seregin describes that “additional MV scaling according to the temporal distance can be applied when converting the reference index to one of the valid EL reference indices.” However, “MV scaling according to the temporal distance” of Seregin only discloses that the MV scaling is according to a single temporal distance, not a ratio between two different temporal distances. Seregin is completely silent with respect to scaling motion vectors “using the determined factor” that is “corresponding to a ratio between the first temporal distance and the second temporal distance.” Indeed, Seregin does not at all disclose or suggest a ratio between two different temporal Appeal 2020-002095 Application 13/545,261 14 distances, much less the specific features of each temporal distance as recited in Claim 21. Id. at 25. Similarly, Appellant argues that this portion of Seregin “does not at all disclose or suggest ‘the second temporal distance’ as recited in claim 21, which is ‘between the corresponding first picture in the first GOP and the first reference picture that is available in the first GOP and absent from the second GOP.’” Id. at 26. In response, the Examiner asserts: In Seregin the motion vector (MV) is originally pointing from the current frame to the missing dropped frame, which corresponds to a first temporal position with a corresponding temporal distance, then it is determined where the next available frame is in accordance with the reference index, which is at a second temporal position with a corresponding temporal distance. The factor by which the MV is scaled by corresponds to the temporal distance difference (i.e., ratio). It is unclear to the Examiner how a MV is scaled, if not by a factor corresponding to a ratio, to achieve a MV pointing to the correct position in an available frame. For instance, if an immediately preceding frame is missing and the subsequent frame is available, then the scaling would be a factor of 2 based on the current position of the MV and the [sic] to be scaled position of the MV. For these reasons, the rejection is maintained. Ans. 49–50. We agree with Appellant that Seregin’s disclosure of motion vector scaling involves a single temporal distance does not teach or suggest the claimed “determining a factor corresponding to a ratio between the first temporal distance and the second temporal distance” as well as the specific requirement that the first picture is in the first GOP and the second reference picture “is preceding to the first reference picture that is available in the first Appeal 2020-002095 Application 13/545,261 15 GOP and absent from the second GOP, the second reference picture available in both the first GOP and the second GOP.” See Appeal Br. 26–27; see also Reply Br. 10. The Examiner has not provided supporting evidence for finding that Seregin teaches or suggests the disputed limitation, and instead has relied on speculative assumptions as to the meaning of the claim terms and the reference teachings in basing the rejection under 35 U.S.C. 103. See Ans. 49–50. “A rejection based on section 103 clearly must rest on a factual basis[.]” In re Warner, 379 F.2d 1011, 1017 (CCPA 1967). “The Patent Office has the initial duty of supplying the factual basis for its rejection. It may not . . . resort to speculation, unfounded assumptions or hindsight reconstruction to supply deficiencies in its factual basis.” Id. Cha In an alternate rejection of claim 21, the Examiner finds Cha discloses the limitations that were found previously in Seregin. Final Act. 27–32 (citing Cha ¶¶ 14, 54). Appellant presents similar arguments that were raised regarding Seregin and asserts that in Cha “neither . . . ‘the distance between the reference frame 60 and the location at which the temporary frame 80 is to be generated’ nor ‘the distance between the reference frame 60 and the inter frame 50’ disclose or suggest ‘the first temporal distance’ as recited in Claim 21.” Appeal Br. 28. Appellant specifically argues that the reference frame disclosed in Cha is different from the recited “second reference picture” which “is available in the first GOP and absent from the second GOP, the second reference picture available in both the first GOP and the second GOP,” as recited in claim 21. Id. (citing Cha ¶¶ 49, 54). Appellant also argues that the reference frame in Cha does not have the Appeal 2020-002095 Application 13/545,261 16 recited “first temporal distance” and “second temporal distance” with the disclosed “inter frame 50” or any frame. Appeal Br. 29. We are persuaded by Appellant’s arguments of Examiner error. We specifically disagree with the Examiner asserting “Appellant has provided no argument directed towards Cha’s teachings of temporally scaling based on a ratio of temporal distances and has provided no supporting evidence that Cha’s temporal scaling cannot be combined with Sarna and Sun’s disclosure of a GOP that is temporally scaled.” See Ans. 54–55. At best, the Examiner’s response improperly shifts the burden of showing the reference teachings to Appellant instead of presenting a prima facie case of obviousness. Accordingly, we agree with Appellant that the Examiner does not provide prima facie support for the rejection. See In re Oetiker, 977 F.2d at 1445. Therefore, we are persuaded that the Examiner errs in rejecting independent claim 21, as well as claim 22 dependent therefrom, as obvious. CONCLUSION For the above reasons, we agree with Appellant that the Examiner’s proposed combination does not teach or suggest the recited features of claims 16, 21, and 23. Therefore, Appellant’s arguments have persuaded us of error in claims 16, 21, and 23, as well as the remaining claims dependent therefrom. See Appeal Br. 31–35 (Claims App.). Appeal 2020-002095 Application 13/545,261 17 DECISION SUMMARY In summary: Claim(s) Rejected 35 U.S.C. § Reference(s)/Basis Affirmed Reversed 21, 22 112(a) Written Description 21, 22 16, 17, 19 103 Sarna, H.264 16, 17, 19 21, 22 103 Sarna, Sun, Seregin 21, 22 23, 27–32, 34–37 103 Sarna, Hannuksela 23, 27–32, 34–37 33 103 Sarna, Hannuksela, Sun, Seregin 33 21, 22 103 Sarna, Sun, Cha 21, 22 Overall Outcome 16, 17, 19, 21–23, 27– 37 REVERSED