Ex Parte Chen et al

Patent Trial and Appeal Board

Dec 8, 2017

13689508 (P.T.A.B. Dec. 8, 2017)

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.
13/689,508 11/29/2012 Ying Chen 1212-207US01/120915U5 8331
15150 7590 12/12/2017
Shumaker & Sieffert, P. A.
1625 Radio Drive, Suite 100
Woodbury, MN 55125
EXAMINER
BRANIFF, CHRISTOPHER
ART UNIT PAPER NUMBER
2484
NOTIFICATION DATE DELIVERY MODE
12/12/2017 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):
pairdocketing @ ssiplaw.com
ocpat_uspto@qualcomm.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte YING CHEN and YE-KUI WANG
Appeal 2017-005523
Application 13/689,5081
Technology Center 2400
Before CARLA M. KRIVAK, HUNG H. BUI, and JON M. JURGOVAN,
Administrative Patent Judges.
BUI, Administrative Patent Judge.
DECISION ON APPEAL
Appellants seek our review under 35 U.S.C. § 134(a) from the
Examiner’s Final Rejection of claims 1-32, which are all the claims pending
in the application. We have jurisdiction under 35 U.S.C. § 6(b).
We AFFIRM-IN-PART.2
1 According to Appellants, the real party in interest is Qualcomm
Incorporated. App. Br. 3.
2 Our Decision refers to Appellants’ Appeal Brief (“App. Br.”) filed March
22, 2016; the Supplemental Appeal Brief (“Supp. Br.”) filed August 4, 2016
submitting only a Claim Appendix listing the correct claims; Reply Brief
(“Reply Br.”) filed February 14, 2017; Examiner’s Answer (“Ans.”) mailed
Appeal 2017-005523
Application 13/689,508
STATEMENT OF THE CASE
Appellants ’ Invention
Appellants’ invention relates to methods and devices for “separately
coding depth and texture components of video data,” and removing a “depth
component... for Multiview Video Coding (MVC) compatible three-
dimensional Video Coding (3DVC).” Title (capitalization altered); Abstract.
Appellants’ invention stores a depth component in a decoded picture buffer,
the depth component pertaining to a view component representing a view in
one time instance, where the view “correspond^] to a different perspective,
or angle, at which corresponding video data of a common scene was
captured.” Spec. ^ 6. Appellants’ invention further analyzes a view
dependency to determine whether the depth component is used for inter­
view prediction, and removes the depth component from the decoded picture
buffer in response to determining that the depth component is not used for
inter-view prediction. Abstract. In one embodiment, Appellants’ invention
removes the depth component of the view component from the decoded
picture buffer without removing a texture component of the same view
component, in response to determining that the depth component is not used
for inter-view prediction. Spec. ^ 6, 142.
Claims 1,9, 17, and 25 are independent. Representative claims 1 and
2 are reproduced below with disputed limitations in italics'.
1. A method for video coding, the method comprising:
storing a depth component in a decoded picture buffer,
analyzing a view dependency to determine whether the
depth component is used for inter-view prediction, wherein the
December 16, 2016; Final Office Action (“Final Act.”) mailed August 27,
2015; and original Specification (“Spec.”), filed November 29, 2012.
2
Appeal 2017-005523
Application 13/689,508
view dependency is associated with the depth component and is
separate from a view dependency for any texture components,
wherein the depth component is associated with a view
component of a view of video data, and wherein a texture
component is also associated with the view component', and
removing the depth component from the decoded picture
buffer in response to determining that the view dependency
associated with the depth component indicates that the depth
component is not used for inter-view prediction.
2. The method of claim 1, wherein removing the depth
component comprises removing the depth component from the
decoded picture buffer without removing the associated texture
component in response to determining that the view
dependency associated with the depth component indicates that
the depth component is not used for inter-view prediction.
Supp. Br. 2-10 (Claims App’x).
Examiner’s Rejections & References
(1) Claims 1, 2, 4, 5, 7-10, 12, 13, 15-18, 20, 21, 23-26, 28, 29,
31, and 32 stand rejected under 35 U.S.C. § 103(a) as being unpatentable
over Jeon et al. (US 2010/0111183 Al; published May 6, 2010; “Jeon”) and
P. Merkle et al. {Multi-View Video Plus Depth Representation and Coding,
IEEE International Conference on Image Processing (ICIP) 201-204 (2007);
“Merkle”). Final Act. 3-7.
(2) Claims 3, 11, 19, and 27 stand rejected under
35 U.S.C. § 103(a) as being unpatentable over Jeon, Merkle, and Pandit et
al. (US 2010/0002762 Al; published Jan. 7, 2010; “Pandit”). Final Act. 7-
8.
(3) Claims 6, 14, 22, and 30 stand rejected under
35 U.S.C. § 103(a) as being unpatentable over Jeon, Merkle, and Y. Chen et
3
Appeal 2017-005523
Application 13/689,508
al. (The Emerging MVC Standard for 3D Video Services, 2009 EURASIP
Journal on Advances in Signal Processing 1-13 (2009); “Chen”). Final Act.
8-10.
Issues on Appeal
Based on Appellants’ arguments, the dispositive issues on appeal are
whether:
(1) the combination of Jeon and Merkle teaches or
suggests (i) “analyzing a view dependency to determine
whether the depth component is used for inter-view prediction,
wherein the view dependency is associated with the depth
component and is separate from a view dependency for any
texture components” wherein “the depth component is
associated with a view component of a view of video data, and
wherein a texture component is also associated with the view
component,” and (ii) “removing the depth component from the
decoded picture buffer in response to determining that the view
dependency associated with the depth component indicates that
the depth component is not used for inter-view prediction,” as
recited in Appellants’ independent claim 1, and similarly
recited in independent claims 9, 17, and 25
App. Br. 6-13; Reply Br. 2-6;
(2) the combination of Jeon and Merkle teaches or
suggests “removing the depth component comprises removing
the depth component from the decoded picture buffer without
removing the associated texture component in response to
determining that the view dependency associated with the depth
component indicates that the depth component is not used for
inter-view prediction,” as recited in Appellants’ dependent
claim 2, and similarly recited in dependent claims 10, 18, and
26
App. Br. 14-15; Reply Br. 7;
4
Appeal 2017-005523
Application 13/689,508
(3) the combination of Jeon, Merkle, and Pandit
teaches or suggests “wherein the depth component does not
belong to a target output view and is a non-reference picture or
a picture marked as ‘unused for reference’,” as recited in
Appellants’ dependent claim 3, and similarly recited in
dependent claims 11, 19, and 27
App. Br. 15-16; Reply Br. 8; and
(4) the combination of Jeon, Merkle, and Chen teaches
or suggests “determining, with a video decoder, that the depth
component and the texture component belong to a target output
view to be output for display; and simultaneously outputting, by
the video decoder, the depth component and the texture
component,” as recited in Appellants’ dependent claim 6, and
similarly recited in dependent claims 14, 22, and 30
App. Br. 16-18; Reply Br. 9-10.
ANALYSIS
Claims 1, 4, 5, 7-9, 12, 13, 15-17, 20, 21,
23-25, 28, 29, 31, and 32
With respect to independent claim 1, the Examiner finds Jeon’s
decoded picture buffer (600) stores a depth component (“depth
information”) for inter-view prediction between camera views with disparity
differences, the depth component being associated with a view component of
a view of video data provided by a camera, and a texture component (e.g., a
color component) also being associated with the view component, as
claimed. Final Act. 4 (citing Jeon]fl| 54, 56, 213-214, 239-240, and Figs. 1
and 14). The Examiner further finds Jeon analyzes the video data’s network
abstraction layer (NAF) having the inter-view dependency information, to
determine whether the depth component is used for inter-view prediction, as
5
Appeal 2017-005523
Application 13/689,508
required by claim 1. Final Act. 4 (citing Jeon ^239, 213); see also Jeon
Figs. 2 and 14. The Examiner additionally finds Jeon’s decoded picture
buffer classifies pictures (corresponding to camera views) based on their use
in performing inter-view prediction, thereby suggesting that views not
needed for inter-view prediction are removed from the decoded picture
buffer. Ans. 11-12 (citing Jeon 59, 239-240). The Examiner concludes
that Jeon teaches “removing the depth component from the decoded picture
buffer in response to determining that the view dependency associated with
the depth component indicates that the depth component is not used for
inter-view prediction,” as claimed. Ans. 11-12; Final Act. 4-5.
To support the conclusion of obviousness, the Examiner relies on
Merkle’s intermediate view synthesized from multiple cameras’ views, for
teaching the claimed “view dependency is associated with the depth
component and is separate from a view dependency for any texture
components.” Ans. 10-11 (citing Merkle § 3.1, Fig. 2, emphasis added);
Final Act. 5 (citing Merkle Fig. 1).
Merkle’s Figures 1 and 2 are reproduced below with additional
markings for illustration.
Fig. I
6
Appeal 2017-005523
Application 13/689,508
Merkle’s Figure 1 illustrates a video plus depth data representation format
for a stereo image pair, the representation including regular 2D color video
and accompanying 8-bit depth-images (or depth maps). Merkle § 2.
Merkle’s Figure 2 illustrates the synthesis of an arbitrary intermediate view
from video plus depth of adjacent camera views from two cameras, for
rendering an intermediate view for any position in between the two cameras.
Merkle § 2.
Appellants dispute the Examiner’s factual findings regarding Jeon and
Merkle. In particular, Appellants argue “Merkle does not in any way
disclose or suggest a view dependency, much less that ‘the view dependency
is associated with the depth component and is separate from a view
dependency for any texture components’”; rather, Merkle merely discloses a
view’s depth component/depth map as a gray scale difference between
frames captured by different cameras, but does not disclose a “view
dependency [that] identifies other views that are predicted from the current
depth component.” App. Br. 9-10; see also Reply Br. 2-5. Appellants
7
Appeal 2017-005523
Application 13/689,508
additionally argue Jeon does not disclose “how a decoded picture buffer is
managed for inter-view prediction,” and does not teach or suggest the
“removing” step recited in claim 1. App. Br. 12-13 (citing Jeon 62, 239-
240); see also Reply Br. 5-6.
We do not find Appellants’ arguments persuasive. Instead, we find
the Examiner has provided a comprehensive response to Appellants’
arguments supported by a preponderance of evidence. Ans. 9-12. As such,
we adopt the Examiner’s findings and explanations provided therein. Id. At
the outset, we note claim terms are given their broadest reasonable
interpretation consistent with the Specification. In re Am. Acad, of Sci. Tech
Ctr., 367 F.3d 1359, 1364 (Fed. Cir. 2004). Under the broadest reasonable
interpretation, claim terms are given their ordinary and customary meaning,
as would be understood by one of ordinary skill in the art in the context of
the entire disclosure. In re Translogic Tech., Inc., 504 F.3d 1249, 1257
(Fed. Cir. 2007).
Appellants’ claim 1 recites, inter alia, a “view,” a “view component,”
a “texture component,” and a “depth component.” Supp. Br. 2 (Claims
App’x). Appellants’ Specification describes a “view component” as “[a]
coded representation of a view in one time instance,” where the “view may
correspond to a different perspective, or angle, at which corresponding video
data of a common scene was captured.” Spec. ^ 6. The Specification
provides that the “view component may contain a depth view component
and a texture view component,” where the texture component may include
“[ljuminance (brightness) and chrominance (color) components” of a left
eye view, and the depth component may be a “depth map . . . computed as a
8
Appeal 2017-005523
Application 13/689,508
difference between the left eye view and [a] right eye view.” Spec. 6, 36,
and 97.
Appellants’ claim 1 further recites “separate” view dependencies, i.e.,
“the view dependency is associated with the depth component and is
separate from a view dependency for any texture components.” Supp. Br. 2
(Claims App’x). Appellants’ Specification does not provide an explicit and
exclusive definition of the claim term “separate” with respect to view
dependencies. Rather, Appellants’ Specification describes “a view
dependency [that] may generally refer to a view from which a view currently
being coded depends” (see Spec. ^ 74), and provides discussion of non­
limiting examples of “separate” view dependencies as follows:
The techniques of this disclosure generally relate to
enabling separate or independent handling of texture
components and depth components of a view when processing
or coding 3D VC video data.
The techniques of this disclosure generally relate to
enabling separate or independent handling of texture
components and depth components of a view when processing
or coding 3D video data.
Because the techniques set forth in this disclosure enable
separate handling of texture and depth components, video
decoder 30 may decode a texture component of a view
component using reference pictures different than those used
when coding a depth component of the same view component
contrary to conventional 3D VC where a view component (and
therefore the depth and texture components of that view
component) was coded using the same reference pictures given
treatment of views in MVC.
Spec. 40, 96, 226 (emphases added).
9
Appeal 2017-005523
Application 13/689,508
Thus, Appellants’ Specification broadly describes depth and texture
view dependencies (for a view component) are “separate'’ when the view
component’s texture component is encoded (and can therefore, be decoded)
“using reference pictures different than those used when coding a depth
component of the same view component.” Spec. ^ 226 (emphasis added).
We, therefore, agree with the Examiner that Merkle’s intermediate
view has separate depth and texture view dependencies commensurate with
the broad description of separate depth and texture view dependencies in
Appellants’ Specification. Ans. 10-11 (citing Merkle § 3.1, Fig. 2). For
example, an intermediate view in Merkle (e.g., Bpicture 3 in Merkle’s
Figure 4, reproduced below) for a position between two cameras (cameras 1
and 3) is a view component with: (i) a depth component predicted from
pictures of cameras 1 and 3 (e.g., from B picture 2 and B picture 4); and (ii)
a texture (color) component predicted from at least one picture (B picture 1)
that is not used for depth prediction. See Merkle §3.1, Figs. 3 and 4. Thus,
we find Merkle’s view component (Bpicture 3) has separate depth and
texture view dependencies because Bpicture i’s texture is predicted from
reference B picture 1 that is different from B pictures 2 and 4 (that are used
for predicting Bpicture 3’s depth).
Merkle’s Figure 4 is reproduced below with additional markings for
illustration.
10
Appeal 2017-005523
Application 13/689,508
(BLACK, HORIZONTAL (RLi>, VERTICAL ARROWS)
ARROWS) p‘g ^
Merkle’s Figure 4 illustrates a multi-view coding structure with hierarchical
B pictures for both temporal prediction and inter-view prediction.
Merkle § 3.1.
Thus, we agree with the Examiner that Merkle discloses a view
dependency determining whether a depth component is used for inter-view
prediction, wherein “the view dependency is associated with the depth
component and is separate from a view dependency for any texture
components” of a view component, as required by claim 1.
Appellants additionally argue Jeon does not disclose the “removing”
step recited in claim 1; rather, Jeon merely discloses memory management
control operations “to manage reference pictures to perform ‘inter-picture
prediction more flexibly’,” but does not disclose “how a decoded picture
buffer is managed for inter-view prediction.” App. Br. 12-13 (citing Jeon
62, 239-240).
11
Appeal 2017-005523
Application 13/689,508
We are not persuaded by Appellants’ arguments, which do not address
the Examiner’s findings regarding what Jeon would have suggested to one of
ordinary skill in the art. Ans. 11-12. Particularly, Jeon discloses a decoded
picture buffer stores reference pictures, which are divided into an inter-view
picture group and a non-inter-view picture group to indicate “that reference
pictures failing to be used in performing inter-view prediction shall not be
used.” Jeon 72; Ans. 11 (citing Jeon 59). In this way, Jeon seeks to
reduce the burden imposed on the decoded picture buffer (DPB). See Jeon
^ 11; Ans. 11-12. We consider the skilled artisan, viewing these teachings
of Jeon, would have recognized the benefit of removing an entire view
component that is not needed for inter-view prediction, thereby reducing the
DPB’s storage burden. Ans. 11-12; Jeon 11.
Additionally, we note claim l’s “removing” step does not preclude
removing the entire view component, that is, removing both depth and
texture components of the view component. As Jeon teaches and suggests
removing an entire view component from the DPB in response to
determining that view component is not used for inter-view prediction, we
agree with the Examiner that Jeon teaches and suggests claim l’s
“removing” step.
Accordingly, Appellants’ arguments have not persuaded us of error in
the Examiner’s rejection of claim 1. As such, we sustain the Examiner’s
obviousness rejection of claim 1 based on Jeon and Merkle, and similarly,
independent claims 9, 17, and 25 for which Appellants provide the same
arguments, and dependent claims 4, 5, 7, 8, 12, 13, 15, 16, 20, 21, 23, 24, 28,
29, 31, and 32, argued for their dependency. App. Br. 14.
12
Appeal 2017-005523
Application 13/689,508
Claims 2, 10, 18, and 26
Claim 2 depends from claim 1, and further recites that “removing the
depth component comprises removing the depth component from the
decoded picture buffer without removing the associated texture component
in response to determining that the view dependency associated with the
depth component indicates that the depth component is not used for inter­
view prediction.” Claim 10, depending from claim 9, claim 18, depending
from claim 17, and claim 26, depending from claim 25, recite similar
limitations.
The Examiner finds Jeon suggests the limitations of claim 2. The
Examiner reasons that
within the well-understood scheme of buffering disclosed by
Jeon—e.g., where decoded information is added and removed
from picture buffers—the use of separate reference picture lists
of view reference information and inter-depth-view reference
information. . . . suggests independent memory management for
texture and depth, with a depth component being removed from
the decoded picture buffer without removing a texture
component, as recited in claim 2.
Ans. 13-14 (emphasis added).
Appellants argue Jeon and Merkle do not teach or suggest separate
handling of a view component’s depth and texture by removing the depth
component from the decoded picture buffer without removing the associated
texture component, as claimed. Reply Br. 7; App. Br. 15. We agree with
Appellants. The cited portions of Jeon suggest, at most, removing an entire
view component from the buffer—that is, removing the view component’s
depth and texture together, as discussed supra with respect to claim 1. Jeon,
however, does not disclose or suggest removing a depth component
13
Appeal 2017-005523
Application 13/689,508
separately from a texture component, as recited in claim 2. Merkle also does
not disclose or suggest removing a depth component separately from a
texture component.
In addition, the Examiner has not provided an adequate reason based
on rational underpinnings to explain why one having ordinary skill in the art
would have been led to modify Jeon or Merkle to remove depth components
without removing associated texture components from a decoded picture
buffer. In re Chaganti, 2014, WL 274514, *4 (Fed. Cir. 2014) (“It is not
enough to say that... to do so would ‘have been obvious to one of ordinary
skill.’ Such circular reasoning is not sufficient-more is needed to sustain an
obviousness rejection.”)
For the reasons set forth above, Appellants have persuaded us of error
in the Examiner’s obviousness rejection of claims 2, 10, 18, and 26.
Accordingly, we do not sustain the Examiner’s rejection of claims 2, 10, 18,
and 26.
Claim 3, 11, 19, and 27
Claim 3 depends from claim 1, and further recites “wherein the depth
component does not belong to a target output view and is a non-reference
picture or a picture marked as ‘unused for reference’.” Claim 11, depending
from claim 9, claim 19, depending from claim 17, and claim 27, depending
from claim 25, recite similar limitations.
The Examiner finds Jeon and Merkle’s teaching of depth and texture
components for view components of a picture, together with Pandit’s
teaching of “marking a reference picture as ‘unused for reference’” for
memory management control, suggest that depth components not used for
14
Appeal 2017-005523
Application 13/689,508
inter-view prediction are marked as “unused for reference.” Final Act. 8
(citing Pandit 3, 39); Ans. 14.
Appellants contend Pandit’s reference pictures “do[] not include
anything that would have been considered as a depth component,” and
because “Pandit simply does not disclose, suggest, or even mention a ‘depth
component’,” Pandit does not cure the deficiencies of Jeon and Merkle.
App. Br. 16 (citing Pandit 21, 39) and Reply Br. 8. However, as
discussed supra, we find Jeon and Merkle are not deficient with respect to
claims 1,9, 17, and 25, and agree with the Examiner that the combination of
Jeon and Merkle teaches and suggests depth and texture components for a
picture or view component.
Additionally, we agree with the Examiner that the skilled artisan
would have recognized the benefit of marking an entire picture—including
the picture’s depth and texture—as “unused for reference,” as taught by
Pandit. Ans. 14. The skilled artisan would have been motivated to do so in
order to reduce the memory’s storage burden with respect to pictures not
needed for inter-view prediction and other coding operations, as taught by
Jeon and Pandit. See Jeon 11, 72; Pandit 2, 4, and 160-161.
As Appellants’ arguments have not persuaded us that the Examiner
erred in rejecting claims 3, 11, 19, and 27, we sustain the Examiner’s
rejection of these claims.
Claims 6, 14, 22, and 30
Claim 6 depends from claim 1, and further recites “determining, with
a video decoder, that the depth component and the texture component belong
to a target output view to be output for display,” and “simultaneously
15
Appeal 2017-005523
Application 13/689,508
outputting, by the video decoder, the depth component and the texture
component.” Claim 14, depending from claim 9, claim 22, depending from
claim 17, and claim 30, depending from claim 25, recite similar limitations.
The Examiner finds Jeon and Merkle’s teaching of depth and texture
components for view components of a picture, together with Chen’s teaching
of simultaneously outputting pictures in a same time instance for display,
suggest that a view component’s depth and texture are simultaneously
outputted, as required by claim 6. Final Act. 9 (citing Chen §§1.1 and 5.2);
Ans. 15.
Appellants contend Chen merely “discuss[es] how 2D views (which
have no depth) are output simultaneously,” and does not disclose “anything
with regard to depth and texture components,” as “Chen is not directed to
3DVC, only MVC [multiview video coding],” which “does not utilize depth
components.” App. Br. 17-18 and Reply Br. 9.
Appellants’ argument is not persuasive. As discussed supra, the
Examiner asserted Jeon and Merkle, not Chen, as teaching views having
depth and texture components. Additionally, we agree with the Examiner
Chen’s simultaneous outputting of views “in the same time instance” for
display (see Chen §§ 1.1, 5.2) suggests simultaneously outputting Merkle’s
camera views that were captured in the same time instance from different
viewing angles (see Merkle Figs. 2 and 4). Ans. 15; Final Act. 10. Since
Merkle’s camera views include depth and texture, simultaneously outputting
Merkle’s camera views will simultaneously output the depth and texture
components of those camera views, thereby teaching and suggesting
“simultaneously outputting, by the video decoder, the depth component and
the texture component,” as recited in claim 6. Ans. 15; Final Act. 10.
16
Appeal 2017-005523
Application 13/689,508
As Appellants’ arguments have not persuaded us the Examiner erred
in rejecting claims 6, 14, 22, and 30, we sustain the Examiner’s rejection of
these claims.
CONCLUSION
On the record before us, we conclude Appellants have demonstrated
the Examiner erred in rejecting claims 2, 10, 18, and 26, but not claims 1, 3-
9, 11-17, 19-25, and 27-32.
DECISION
As such, we AFFIRM the Examiner’s Final Rejection of claims 1, 3-
9, 11-17, 19-25, and 27-32. However, we REVERSE the Examiner’s Final
Rejection of claims 2, 10, 18, and 26.
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a)(l)(iv).
AFFIRMED-IN-PART
17