Nvidia Corporation

Patent Trials and Appeals Board

May 18, 2021

2020000063 (P.T.A.B. May. 18, 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.
15/043,978 02/15/2016 Thrinadh Kottana 14-BG-0317-US01 7439
102469 7590 05/18/2021
PARKER JUSTISS, P.C./Nvidia
14241 DALLAS PARKWAY
SUITE 620
DALLAS, TX 75254
EXAMINER
HOSSAIN, FARZANA E
ART UNIT PAPER NUMBER
2482
NOTIFICATION DATE DELIVERY MODE
05/18/2021 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):
docket@pj-iplaw.com
PTOL-90A (Rev. 04/07)

UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte THRINADH KOTTANA, VINAYAK PORE,
CHIRAYU GARG, and SOUMEN KUMAR DEY

Appeal 2020-000063
Application 15/043,978
Technology Center 2400
Before MAHSHID D. SAADAT, RICHARD M. LEBOVITZ, and
MATTHEW J. MCNEILL, Administrative Patent Judges.

LEBOVITZ, Administrative Patent Judge.

DECISION ON APPEAL
The Examiner rejected claims 1–20 under 35 U.S.C. § 102 as
anticipated and under 35 U.S.C. § 103 as obvious. Pursuant to 35 U.S.C.
§ 134(a), Appellant1 appeals from the Examiner’s decision to reject the
claims. 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. Appellant identifies the real party in interest as Nvidia
Corporation. Appeal Br. 3.
Appeal 2020-000063
Application 15/043,978

2
STATEMENT OF THE CASE
The Examiner rejected claims 1–20 in the Final Office Action (“Final
Act.”) as follows:
Claims 1–3, 5, 6, 8–12, and 14–19 under 35 U.S.C. § 102(a)(1) as
anticipated by Tokumo et al. (US 2009/0262837 A1, published Oct. 22,
2009) (“Tokumo”). Final Act. 4.
Claims 4, 7, 13, and 20 under 35 U.S.C. § 103 as obvious in view of
Tokumo and Nguyen et al. (US 2010/0150232 A1, published Jun. 17, 2010)
(“Nguyen”). Final Act. 9.
Independent claim 1 is representative and copied below:
1. A video viewing device, comprising:
a screen;
a decoder configured to decode a data frame received in a
bitstream from a transmitter to provide a decoded data frame;
and
an error concealer configured to discard said decoded
data frame when a content complexity of said decoded data
frame is higher than a content complexity threshold and select
said decoded data frame for display on said screen when said
content complexity of said decoded data frame is lower than
said content complexity threshold.

REJECTIONS
Anticipation
The Examiner rejected claims 1–3, 5, 6, 8–12, and 14–19 as
anticipated by Tokumo. The Examiner found that Tokumo describes a video
viewing device with a screen, decoder, and error concealer, the same three
elements required by independent claim 1. Final Act. 4–5. The Examiner
further found that the error concealer described by Tokumo uses a
complexity threshold to determine whether to discard or select a frame for
Appeal 2020-000063
Application 15/043,978

3
display, citing paragraph 75 of Tokumo for its discussion of using a
threshold to determine when to display a frame. Final Act. 5. A “content
complexity threshold” is required by the last clause of claim 1. The
Examiner also generally cited paragraphs 62–75 in Tokumo to support the
finding that Tokumo anticipates claim 1. Id.
Appellant argues that paragraph 75 of Tokumo describes discarding
frames based on the number of error blocks in a frame. Appeal Br. 6.
Appellant contends that “the number of error macro blocks is not one of the
frame statistics used to determine the content complexity (see, e.g., par. 19
of the specification).” Id. at 6–7. Appellant further argues that the number of
error macro blocks in a frame “cannot be reasonably interpreted as a statistic
related to the content complexity because the content complexity can vary
irrespective of the number of error macro blocks.” Id. at 7. Rather, Appellant
argues that error blocks are used by Tokumo to determine image quality. Id.
The Examiner responds that while “error macro blocks,” as described
by Tokumo in paragraph 75, are not listed in the Specification as a frame
statistic indicative of content complexity, “limitations from the specification
are not read into the claims.” Ans. 9–10. The Examiner also responded that
Appellant “did not indicate what a content complexity is in the claim.”
Ans. 10.
The issue before us is whether Tokumo describes using a content
complexity threshold to determine when to discard and when to select a
decoded data frame of a bitstream, as required by the last clause of claim 1.
We begin with claim interpretation because before a claim can be
compared to the prior art, its scope must be understood. During patent
examination proceedings, claim terms are given “the broadest reasonable
Appeal 2020-000063
Application 15/043,978

4
meaning . . . in their ordinary usage as they would be understood by one of
ordinary skill in the art, taking into account whatever enlightenment by way
of definitions or otherwise that may be afforded by the written description
contained in the applicant’s specification.” In re Morris, 127 F.3d 1048,
1054 (Fed. Cir. 1997).
Claim 1 is directed to a video viewing device with a screen, a decoder,
and an error concealer. As explained in the Specification, in streaming
games over the cloud for playing on a remote device, the game output is in
the form of a video stream, which is typically encoded into packets (the
“bitstream” in claim 1). Spec. ¶ 2. The packets are transferred to the client
device for viewing. Id. The claimed “screen” is the device component on
which the video stream is viewed. The encoded bitstream received by the
device must be decoded into frames for it be viewed on the screen. This step
is performed by the claimed “decoder.”
The Specification explains that packets may be lost during streaming
which results in “stutters during game play and severe corruption.” Spec.
¶ 4. To compensate for these losses, error concealment methods are used. Id.
The claimed “error concealer” performs this function on the decoded
bitstream.
The Specification further explains, in its description of the invention,
that “the subjective video quality of the final outcome of the conventional
[error] concealing methods at the client/receiver is highly dependent on the
complexity of the video/game content” and can result in frame corruption of
not only the received frame but in subsequent frames. Spec. ¶ 16. The
Specification states that “the existing methods for hiding streaming errors
fail to consider the complexity of the streaming content.” Spec. ¶ 17. To
Appeal 2020-000063
Application 15/043,978

5
address this problem, the Specification discloses that it “provides a quality
aware error concealment (QEC) scheme that understands the complexity of
the video or game content and intelligently determines whether a frame of
the video or game content should be dropped or can be used as a reference
frame.” Spec. ¶ 18. This solution described in the Specification is performed
by the claimed “error concealer” which is “configured to discard” the
decoded data frame “when a content complexity of said decoded data frame
is higher than a content complexity threshold” and “select said decoded data
frame for display on said screen when said content complexity of said
decoded data frame is lower than said content complexity threshold.” Thus,
the claimed method uses the “content complexity” of the decoded bitstream
frames to determine when to drop a frame or when to select a frame for
display on a screen. By dropping or discarding a frame when the complexity
is higher than a threshold value, the Specification states that the claimed
method “reduces frame corruption and error propagation while also reducing
frame stutters of the streaming video or game content.” Spec. ¶ 18.
The Specification describes complexity as follows:
In various embodiments disclosed herein, the complexity of
streaming content is understood by examining statistics from the
bit stream of the streaming content and decoded frames of the
streaming content. Complexity can be estimated using frame
statistics such as average bits per macroblock, motion
information, number of inter/intra macro-blocks in the frame,
etc. Frame statistics used for estimating complexity can be
generated during the decoding process at the client side or can be
sent from the server. One skilled in the art will understand that
the above noted statistics are not exhaustive; other statistics can
be incorporated into the QEC logic disclosed herein to estimate
complexity.
Spec. ¶ 19.
Appeal 2020-000063
Application 15/043,978

6
The error concealer 230 is further configured to determine the
complexity based on a frame statistic or statistics of the data
frame. In one embodiment, the frame statistics include at least
one of or a combination of the following core statistics: average
bits per macro-block in the data frame, motion information
(horizontal or vertical), number of inter macro-blocks in the data
frame, number of intra macro-blocks in the data frame, bitrate
(BR) and the number of skipped macro-blocks. Additional
statistics, such as PSNR [peak signal-to-noise ratio] between
consecutive data frames of the data frame, can also be used to
determine complexity.
Spec. ¶ 37.
We agree with the Examiner that limitations from the Specification
are not ordinarily imported into the claims. Sjolund v. Musland, 847 F.2d
1573, 1581 (Fed. Cir. 1988); In re Van Geuns, 988 F.2d 1181, 1184 (Fed.
Cir. 1993). However, “[c]laims must be read in view of the specification, of
which they are a part.” Markman v. Westview Instruments, Inc., 52 F.3d 967,
979 (Fed. Cir. 1995) (en banc). A proper claim construction analysis
endeavors to assign a meaning to a disputed claim term “that corresponds
with . . . how the inventor describes his invention in the specification.” In re
Smith Int’l, Inc., 871 F.3d 1375, 1383 (Fed. Cir. 2017).
In this case, the Specification describes “complexity” as being
determined based on frame statistics of the decoded bit stream. Spec. ¶¶ 19,
37. Various examples of the frame statistics are disclosed in the
Specification, including “average bits per macro-block in the data frame,
motion information (horizontal or vertical), number of inter macro-blocks in
the data frame, number of intra macro-blocks in the data frame, bitrate (BR)
and the number of skipped macro-blocks.” Spec. ¶ 37. The Specification
explains that mentioned statistics are not “exhaustive” and that “[a]dditional
statistics . . . can also be used to determine complexity.” Spec. ¶¶ 19, 37.
Appeal 2020-000063
Application 15/043,978

7
Nonetheless, from the description of frame statistics in the Specification, it
can be discerned that “content complexity” corresponds to certain attributes
of the frame, such as macro-block and motion information.
We next turn to the dispositive issue in this rejection of whether the
error-blocks described in Tokumo are indicative of content complexity as we
interpret this term in light of the Specification. The Examiner cited
paragraph 75 of Tokumo as disclosing content complexity and using it to
determine when to discard and when to select a frame for viewing.
Paragraph 75 of Tokumo is reproduced below:
If the number nE of the error macro blocks is greater than a
predetermined threshold T2 (nE> T2), the frame may be
discarded without performing the concealment. This can prevent
the image quality from deteriorating due to displaying a frame
including an extremely large number of errors.
Tokumo ¶ 75.
Tokumo does not define “error macro blocks.” Tokumo shows a
decoded frame having 99 blocks in total, where the blocks in the frame are
categorized as intra-macro blocks, inter-macro blocks, and error macro
blocks. Tokumo ¶¶ 65–69; Fig. 5. The inter- and intra-blocks, as indicated
by Appellant, are used by Tokumo to determine an “image for
concealment.” Id. ¶¶ 72, 74, 76. The Examiner did not cite them as
indicative of content complexity for determining when to discard or select a
frame. The error macro blocks are used to determine when to discard a
frame. Id. ¶ 75. The Examiner, however, did not explain how “error” blocks
are a measure of “complexity,” when complexity is described in the
Specification with respect to attributes of the frame, such as “average bits
per macro-block in the data frame, motion information (horizontal or
vertical), number of inter macro-blocks in the data frame, number of intra
Appeal 2020-000063
Application 15/043,978

8
macro-blocks in the data frame, bitrate (BR) and the number of skipped
macro-blocks.” Spec. ¶ 37. Although the Examiner characterized the number
of error blocks in Tokumo as part of the “statistics” related to a frame (see
Ans. 10), the Examiner did not consider the type of statistics associated with
a frame complexity, as described in Appellant’s disclosure cited above. The
Specification discloses that “existing methods for hiding streaming errors
fail to consider the complexity of the streaming content.” Spec. ¶ 17; see
supra. 4–5. Thus, the Specification distinguishes between errors and
complexity. The Examiner did not address this difference in how the terms
are used in the Specification.
“An examiner bears the initial burden of presenting a prima facie case
of obviousness.” In re Huai-Hung Kao, 639 F.3d 1057, 1066 (Fed. Cir.
2011). Here, the claim requires discarding a decoded frame “when a content
complexity of said decoded data frame is higher than a content complexity
threshold” and selecting “said decoded data frame for display on said screen
when said content complexity of said decoded data frame is lower than said
content complexity threshold.” The claim therefore uses “content
complexity” to determine when to discard a frame and when to select a
frame. The Examiner did not adequately explain how the “error macro
block” described in Tokumo represents content complexity as required by
the claim, when the inventors used “content complexity” to describe an
attribute of a frame (Spec. ¶¶ 19, 37) and used the term “error” is a different
context (Spec. ¶ 17). See Smith, 871 F.3d at 1383.
A claim is anticipated under 35 U.S.C. § 102 when publication
“disclose[s] all elements of the claim within the four corners of the
document” and “arranged as in the claim.” Net MoneyIN, Inc. v. VeriSign,
Appeal 2020-000063
Application 15/043,978

9
Inc., 545 F.3d 1359, 1369 (Fed. Cir. 2008). Because the Examiner did not
establish that error macro blocks is an attribute of complexity as required by
the claim 1, we are compelled to reverse the anticipation rejection.
Independent claims 8 and 15 also require complexity threshold to discard a
frame and therefore are reversed for the same reasons. Dependent claims 2,
3, 5, 6, 9–12, 14, and 16–19 incorporate the limitations of the independent
claims and are reversed for the same reasons.

Obviousness
The Examiner further cited Nguyen for limitations in dependent
claims 4, 7, 13, and 20. The Examiner found that Nguyen discloses an error
concealer that receives complexity information from a transmitter, such as
frame statistics. Final Act. 9. However, the Examiner did not explain, based
on the combination of Tokumo and Nguyen, why it would have been
obvious to one of ordinary skill in the art to use a complexity threshold to
determine when to discard and select a frame as required by the claim.
Accordingly, the obviousness rejection of claims 4, 7, 13, and 20 is reversed.

CONCLUSION
In summary:
Claims
Rejected
35 U.S.C.
§
Reference(s)/Basis Affirmed Reversed
1–3, 5, 6, 8–
12, 14–19
102 Tokumo 1–3, 5, 6, 8–
12, 14–19
4, 7, 13, 20 103 Tokumo, Nguyen 4, 7, 13, 20
Overall
Outcome
1–20

REVERSED