Trials@uspto.gov
`571-272-7822
`
`
`
`
`Paper No. 40
`Date: August 18, 2020
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`UNIFIED PATENTS, LLC,
`Petitioner,
`
`v.
`
`VELOS MEDIA, LLC,
`Patent Owner.
`____________
`
`IPR2019-00757
`Patent 9,930,365 B2
`____________
`
`Record of Oral Hearing
`Held: June 16, 2020
`____________
`
`
`
`
`Before MONICA S. ULLAGADDI, JASON W. MELVIN, and
`AARON W. MOORE, Administrative Patent Judges.
`
`
`
`
`
`
`
`
`
`
`
`
`571-272-7822
`
`
`
`
`Paper No. 40
`Date: August 18, 2020
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`UNIFIED PATENTS, LLC,
`Petitioner,
`
`v.
`
`VELOS MEDIA, LLC,
`Patent Owner.
`____________
`
`IPR2019-00757
`Patent 9,930,365 B2
`____________
`
`Record of Oral Hearing
`Held: June 16, 2020
`____________
`
`
`
`
`Before MONICA S. ULLAGADDI, JASON W. MELVIN, and
`AARON W. MOORE, Administrative Patent Judges.
`
`
`
`
`
`
`
`
`
`
`
`
`
`IPR2019-00757
`Patent 9,930,365 B2
`
`
`
`APPEARANCES:
`
`ON BEHALF OF THE PETITIONER:
`
`
`ERIC A. BURESH, ESQ.
`7015 College Blvd.
`Suite 700
`Overland Park, KS 66211
`
`
`
`ON BEHALF OF THE PATENT OWNER:
`
`
`BARRY J. BUMGARDNER, ESQ.
`CHRIS BRANNIGAN, ESQ.
`KENNY MCCLURE, ESQ.
`Nelson Bumgardner Albritton, P.C.
`3131 West 7th St., Suite 300
`Fort Worth, TX 76107
`
`
`
`
`The above-entitled matter came on for hearing on Tuesday, June 16,
`
`2020, commencing at 9:00 a.m. EDT, by video/telephone.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`2
`
`
`
`IPR2019-00757
`Patent 9,930,365 B2
`
`
`P R O C E E D I N G S
`- - - - -
`JUDGE ULLAGADDI: Good morning and welcome to the Patent
`Trial Appeal Board. We are here today for oral arguments and interparty
`reviews matter number 2019-00757. A case in which Unified Patents is the
`Petitioner and Velos Media is the Patent Owner. At issue is U.S. Patent
`number 9,930,365.
`Your panel for the hearing today includes myself, Judge Ullagaddi,
`Judge Melvin and Judge Moore. I would like to start by getting the
`appearances of counsel. Who do we have on behalf of Petitioner?
`MR. BURESH: Your Honor, this is Eric Buresh of Erise IP on behalf
`of Unified Patent, LLC, and with me and also who will be appearing is
`David Cavanaugh of WilmerHale on behalf of Unified Patent.
`JUDGE ULLAGADDI: Thank you, Mr. Buresh. Who do we have on
`behalf of Patent Owner?
`MR. BUMGARDNER: Good morning, Your Honor. This is Barry
`Bumgardner here on behalf of Patent Owner with me here physically and
`present in my office is Chris Brannigan, another attorney of record who will
`be discussing some real party interest issues. Also, present with me in the
`office is Kenny McClure in-house counsel with Velos and then also on the
`dial-in line is another in-house attorney with Velos: Rory Litten.
`JUDGE ULLAGADDI: Thank you, Mr. Bumgardner. And thank
`you all for joining us today. I've got a few administrative details about the
`procedure and format of the hearing that I would like to go over before we
`get started. When you are working through your demonstrative slides,
`please make sure that you reference the demonstrative slides that you are
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`
`
`3
`
`
`
`IPR2019-00757
`Patent 9,930,365 B2
`
`using because we can't see them.
`Each party will have 60 minutes total to argue their case. As Patent
`Owner has requested a bifurcated trial, we are going to start with the
`nonconfidential portion of trial. We'll first hear from Petitioner. Petitioner,
`you will present your arguments related to patentability.
`Patent Owner, you will then be permitted to present your argument.
`Petitioner would you like to reserve any time for rebuttal today?
`MR. BURESH: As to the ethical side of the nonconfidential side,
`Your Honor, I would like to reserve 15 minutes.
`JUDGE ULLAGADDI: Fifteen minutes reserved on rebuttal; okay.
`MR. BURESH: And just to clarify, Your Honor, I intend to go 35
`minutes in my opening and approximately 15 minutes in rebuttal and then
`we'll reserve approximately 10 minutes for our RPI rebuttal.
`JUDGE ULLAGADDI: Thank you. When you are ready you may
`begin.
`MR. BURESH: Thank you, Your Honor. I am going to be working
`with Petitioner's demonstrative slides which you should have in front of you,
`starting at slide number 2 just marked B. I'm going to be discussing first
`partitioning encoding and then I'm going to provide a little background of
`the 365 patent, the challenged patent.
`Then I will discuss how partitioning occurs in the primary prior art
`reference -- and we'll address the key prime limitations in that portion - there
`will be a short portion on Novotny's A and B syntax element called for by
`the challenged claims - and then I will be wrapping up.
`Moving to slide number 3 in reviewing coding, Your Honors, the
`purpose of partitioning is really to simplify it down as much as I can, bigger
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`
`
`4
`
`
`
`IPR2019-00757
`Patent 9,930,365 B2
`
`blocks are better. They're more efficient, but when video images or picture
`images have granule area and particularly are in demand and particular
`features that will be changing from image to image, it becomes necessary to
`utilize smaller blocks in order to capture the fine nature of those edge pieces.
`So, on slide number three, we have an example of what might be an
`ideal partitioning where you see background information is depicted and
`captured through larger blocks and then when you get to edge features where
`there will be more change from image to image, you go or work your way
`down to smaller blocks, which would be partitioned off the larger blocks.
`This is depicted in figure 12 of the 365 patent where you see the idea
`of larger blocks end-to-end and working or segmenting down to smaller
`blocks within that larger block. Also, depicted on figure 12 is the types of
`decoding modes that might be used for the respective blocks and smaller
`blocks within those blocks.
`That is the underlying purpose for partitioning. At this timeframe,
`Your Honors, virtually all decoding and coding is image including the main
`encoder and decoder we utilize - what is called rate distortion analysis - to
`determine the proper sizing of what is desirable sizing of blocks and what
`rate distortion analysis is effectively a cost benefit where you learn how
`much quality of video are you losing in order to go with a larger block that is
`marked “essential patent.”
`So, you're looking at both the quality size and the efficiency size and
`using an analysis. There are a number of different analyses that can be used
`by using two-piece analysis to find the best sizing for your blocks. And if
`we go to slide number 6 after slide number 5 outline, you will see in the 365
`patent and before I go any further on slide number 6, many of the Patent
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`
`
`5
`
`
`
`IPR2019-00757
`Patent 9,930,365 B2
`
`Owner arguments in this case are what I would call “higher arguments”
`where the patent owner is attempting to distinguish in a broad brush fashion
`between prior referenced Kalker and the 365 patent.
`And one of the impressions, at least I received, when the patent owner
`responded and also the patent owner served, was a suggestion that the 365,
`and this is drawn from the fact that the claims of the 365 in this particular
`application are all directed to the decoder. But the suggestion was that the
`365 has potentially very intelligent decoders and was able to make
`segmentation additions on its own, separate and apart from the encoders.
`And one of the purposes of these next few slides is just to debunk that.
`The 365 patent is the same basic structure as all systems in the timeframe
`where the encoder is making the decision. The encoder is intact information
`to the decoder and the decoder is simply reversing the process of the
`encoder. And that reversing of the decisions that were made by the encoder
`is what the 365 patent refers to as partitioning.
`So, the partitioning decision is made on the encoder side. Syntax
`information is transmitted to the decoder and the decoder reverses using that
`syntax information to recreate an image and that uses data in itself
`partitioning at the decoder. And we'll see that -
`JUDGE ULLAGADDI: May I ask a question about that?
`MR. BURESH: Sure.
`JUDGE ULLAGADDI: If the encoder side is doing the partitioning
`then why would the decoder side also partition? Wouldn't it be putting
`everything back together?
`MR. BURESH: Yeah, and that's the point, Your Honor. The decoder
`side in the act of - as the words you use "putting everything back together" it
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`
`
`6
`
`
`
`IPR2019-00757
`Patent 9,930,365 B2
`
`is recreating a segmentation after a partition map and that act of taking the
`video image that in its original data stream is that the whole image and
`bringing it in through the decoder and breaking it up into blocks that will be
`processed on a block -
`That's efficient of the image blocks is the partitioning on the decoder
`side. Now, as I said, the decoder is not making the decision about what the
`block size will be, but when the decoder devised the image into blocks the
`process then is exactly the same way and in exactly the same block sizes as
`the encoder is partitioning. And that is what is described in the 365 patent.
`JUDGE ULLAGADDI: Thank you.
`MR. BURESH: And if we go to slide 6, the main point here - a
`couple of points, one the encoder in the 365 patent is determining the
`different partitioning levels for the macro block. Just like in Kalker it is the
`encoder that is making the determination or the partitioning in the 365 patent
`and that determination that the partition levels as well as decoding modes to
`apply to the partition is based on a rate distortion analysis.
`I highlight that just to show the similarities between the 365 and
`Kalker which we'll see in a minute, uses exactly that same rate distortion
`analysis to make the segmentation decision at the encoder. Turning to slide
`number seven, I'm going to go quickly through these next two.
`Again, just to show a correlation between the 365 patent and Kalker,
`the 365 patent on slide number 7 is talking about the encoder selecting the
`syntax or excuse me block base syntax to use based on the largest block.
`We will see also the Kalker's that once it's blocked at least in the
`embodiment we are relying upon in the partition is also using the largest
`block syntax.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`
`
`7
`
`
`
`IPR2019-00757
`Patent 9,930,365 B2
`
`
`And again, the video decoder turning to slide number 8, is receiving
`that syntax information from the video encoder and is similarly using it to
`select the same block-based syntax information based upon the largest
`blocks. Again, we see between slides 7 and 8 that the decision made at the
`encoder to the decoder to recreate.
`And I'm going to get to the little more granular on that as we go
`because I think it's important that it's not just in the 365 patent turning to
`slide nine, it is not just a coded unit syntax information. There is syntax
`information in the 365 that is sent from the encoder to the decoder that
`actually shows the partitioning of individual blocks as well.
`So, you have the largest block size and smallest block size that is
`transmitted in the 365 patent, but you also have partitioning for the
`individual blocks. Turning to slide 10, we're going to see what the syntax
`information is and what it reflects from slide 9. In slide 10, again, you're
`looking at the video encoder side and what it is generating in a second block
`on slide 10.
`The video encoder is generating what type syntax information that
`indicates the size of a block and identifies the partitioning in the encoding
`mode used to encode the partition. So, we see again, the encoder in the 365
`patent is making the partitioning decision. It is generating syntax
`information to reflect those partitioning decisions and it is transmitting those
`partitioning syntax - that particular syntax information to the decoder so that
`the decoder can itself reverse the process which the 365 calls partitioning.
`And while the 365 does not go into any great detail about the
`decoders, the vast majority of the description in the 365 relates to the
`encoder, but if we turn to slide 11 and look at the pending claim three which
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`
`
`8
`
`
`
`IPR2019-00757
`Patent 9,930,365 B2
`
`depends on challenge claim one we see that the decoder receives and further
`comprising decoding one or more syntax elements representative of
`partitioning of the current block. So, in claim one we have the concept of
`the first syntax element that is the smallest block in the encoded unit, a
`second syntax element that is the largest block in the encoded unit and the
`point here is there is additional syntax information that is actually
`representing the partitioning for the current block. So, between the largest
`and the smallest how a particular block is partitioned is based upon
`information sent from the decoder - encoder and received by the decoder.
`Any questions on that, Your Honors?
`JUDGE ULLAGADDI: No.
`MR. BURESH: One more point before we leave the 365 patent on
`slide number 12, this is responsive to the arguments by Patent Owner that
`Kalker uses the words "blocks and smaller blocks" - use the word
`“subblocks” in Kalker and the point of slide 12 is simply to note that within
`the 365 patent the notion of a “subblock” is also a block. You will always
`have blocks. You will have larger blocks you can partition into smaller
`blocks.
`The smaller blocks can be called subblocks, but those subblocks can
`just be called smaller blocks. The nomenclature there, there is nothing
`special about subblocks except that they are a smaller block within a larger
`block.
`Skipping over slide 13 which simply indicates that I'm moving onto
`Kalker now. Slide number 14 the disclosing in Kalker again, the basics if
`you just look at figure one, it's a very easy kind of reference here in Kalker
`to get the basic framework where you have a second circuit that generates a
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`
`
`9
`
`
`
`IPR2019-00757
`Patent 9,930,365 B2
`
`second of video images into blocks.
`You have a map encoder that generates your segmentation map based
`upon the segmentation decided by the segmentation circuit. That
`information is transmitted to a decoder which again reverses the process and
`gives you a segmentation map that can be applied to inverse the operation on
`the encoder side.
`JUDGE ULLAGADDI: I think I'd like to come back to that point. If
`the decoder side is inversing as you say operations made on the encoder
`side, what is the inverse of partitioning?
`MR. BURESH: So, if you look at figure number1 in Kalker which is
`tighter on the page in the figures, but in figure number 1 you see the concept
`of the -- coming out of the segmentation circuit. That is -- what the
`segmentation circuit does is it takes an input whole video image and it
`determines which is on slide number 14 actually based upon that rate
`distortion analysis the same one in the 365 patent, it determines at the
`segmentation circuit the optimal block size.
`That optimal block size is then encoded by the map encoders. So, you
`have your block sizes that are desirable, the map encoder will generate
`syntax information to reflect that in a segmentation map. When the map
`decoder receives that information that encoded segmentation map what it
`does is takes -- it generates a new segment or it generates the same
`segmentation map reflected by apps coming out of the map decoder and
`applies that and sends the S to the inverse transform circuits.
`So, what the inverse transform circuit is doing is taking the incoming
`video information that comes through the transform decoder. It is
`segmenting that incoming video information into the identical block sizes
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`
`
`10
`
`
`
`IPR2019-00757
`Patent 9,930,365 B2
`
`that were encoded at the transform coder of the encoder and then using those
`block sizes which correspond identically to the ones that are incoming from
`the encoder, it decodes.
`So, the segmentation on the reverse side is taking your incoming
`stream of video information and dividing it into the identical block sizes that
`were used during the encoding process. So, when you have an incoming
`video stream you're going to -- in the Kalker example it is your first coded
`syntax is a three, you're going to take an incoming video stream and you're
`going to take the first 16 by 16 bit and what the segmentation is taking your
`incoming video stream and taking it into those segments, 16 by 16.
`Now, the stream is continuing to come in and I may take the next one
`in an 8 by 8 segment and the one after that in a 4 by 4 segment or four sets
`of 4 by 4 segments. So, as the incoming video stream is being processed it's
`being segmented into chunks in a very and I apologize for the nontechnical
`term, but it is being segmented into chunks that will be processed in those
`chunks or blocks that correspond to the blocks or segmentation that was
`provided by the encoder.
`JUDGE ULLAGADDI: It is the Petitioner's --
`JUDGE MELVIN: Don't those S values represent just the size of a
`particular block and not a minimum or maximum size of the blocks in the
`image?
`MR. BURESH: The particular value as the stream of S's the 3’s, 2’ss
`and the 1’s are coming into the map encoder and those S values do represent
`the size of a particular block. In the petition we were very clear to
`distinguish, if you look at, for example, note 5 on page 29 of our petition.
`The same distinctions are drawn in the expert declaration.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`
`
`11
`
`
`
`IPR2019-00757
`Patent 9,930,365 B2
`
`
`The value of S represents an individual block, but it is the syntax
`information that establishes the relationship between S and a particular block
`size and in the example in Kalker, the information that establishes that S
`equals 16 by 16, I'm sorry S equals three equals 16 by 16, is the syntax
`element they were pointing to for the second syntax element.
`The syntax information that is sent to the decoder to indicate that S
`equals one equals 4 by 4, is the first syntax element representing the smallest
`block.
`And that assignment information and is used to in Kalker generates a
`grid of the largest block size from which the system then operates, that is the
`syntax element that satisfies or syntax information that satisfies the first and
`second syntax elements. As I walk through this I can get into more detail. I
`understand that's based upon the questions and the institution decision, I
`understand that's a question and I'll continue to address that.
`Slide 14, quickly, just reiterates the segmentation is done at the
`encoder in Kalker by a segmentation circuit which is implemented into a
`segmentation map by the map encoder, it's based on a rate distortion
`analysis. The same basic analysis within the 365 patent.
`The video encoder scans in slide 15, the video encoder will scan based
`on the largest block size in the encoded unit. So, we note from this that that
`largest block size just like in the 365 is important and the way the largest
`block size is used if we turn to slide 16, and I want to note here, Your
`Honors, there are two embodiments in Kalker.
`And that the two embodiments are where Kalker will establish a grid,
`a segmentation grid based upon the smallest block size and then scan that
`grid based on the smallest block size and we are not relying on that
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`
`
`12
`
`
`
`IPR2019-00757
`Patent 9,930,365 B2
`
`embodiment. I think you could argue whether it maps to these claims but
`we're not relying on that because there is a clear embodiment where you
`have Kalker generating the segmentation grid based on the largest block size
`and that is what I'm going to walkthrough.
`It's introduced here on slide 16 where you have three lines down and
`the quote first the top left of 16 by 16 block is analyzed and then the final
`sentence then the next 16 by 16 block is analyzed. So, from that we see that
`we're moving through the segmentation grid on a largest block size scale.
`We're moving 16 byblock by 16 by 16 block and that's further described on
`slide 17.
`Within those larger blocks of 16 by 16 as we proceed through the
`segmentation map you will also see this notion of dividing those larger
`starting blocks inset into smaller blocks, but for present purposes, what I'm
`trying to get across here is that the largest block size S equals three which
`represents 16 by 16, is used by the decoder to generate the segmentation
`map with a grid of the largest block size so that the processing can then
`occur on that largest block size grid from largest block to largest block to
`largest block if you proceed through a coded unit.
`JUDGE MELVIN: So, where does Kalker determine that the current
`block as its starting is equal to the maximum size?
`MR. BURESH: Yes. So, that will -- I have that claim on slide 21, if
`you turn there. Just in passing, as we go past slide 20 and I will answer your
`question. The slide 20 is just kind of to close the loop here and show that
`the segmentation information that is generated at the encoder is in fact
`transmitted to the decoder.
`It says here in Kalker, it must be transmitted to the receiver to allow it
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`
`
`13
`
`
`
`IPR2019-00757
`Patent 9,930,365 B2
`
`to apply a backward transform to the same block as was used by the encoder.
`And that is just a preview. I include that to address the demonstrative that
`Patent Owner has brought forth in the reply which suggested that there are
`essentially different block sizes that could be somehow generated by the
`decoder itself.
`Slide 21 is determining that a block or a plurality of blocks and a
`sequence of pictures has a starting size equal to the maximum size using a
`second syntax element. That is the claim limitation Your Honor was
`referring to. And again, the way this operates in Kalker is that the largest
`block size in a particular coded unit is transmitted from the encoder to the
`decoder, which is the second syntax element.
`That second syntax element in the second embodiment in Kalker is
`used to establish a grid size at the decoder using the largest block size. And
`on slide 22, we see that the segmentation map is scanned on the basis of the
`largest block size. Now, in Kalker, if we turn back to -- go back to slide 18.
`That grid establishes the -- is used to determine the starting block size
`of the block. Now, if the starting block size in the claim limitation we're
`looking at here, is not necessarily the block size. It is just the starting block
`size. It may be the block size if the block that you're looking at, the current
`block, is at the largest block size.
`So, for instance, the upper left-hand block as you start your analysis in
`Kalker you assume it's a 16 byblock, so that is the starting size for your
`current block. You then assess the syntax information for that block and in
`this example the syntax information for the block is three. So, that current
`block is at the largest block size and no further syntax information is
`necessary.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`
`
`14
`
`
`
`IPR2019-00757
`Patent 9,930,365 B2
`
`
`Then if you proceed to the next top right 16 by 16 block, so the
`starting size of that block was determined based upon the coded unit syntax
`information, the second syntax element that the starting size of that block
`would be 16 by 16. That block, however, the syntax information related to
`that particular block indicates that there is further sub partitioning.
`So, we need to go through and scan that block to determine what the
`subblocks are within that starting size of 16 by 16. So, the starting size for
`the current top right block was 16 by 16 based upon the coded unit syntax
`information establishing the grid for that starting size to be 16 by 16.
`From there, so you have current block has a starting size. The starting
`size of 16 by 16 and from there you go forth to a further partition based upon
`the syntax information indicating the partitioning of that particular block.
`So, Your Honor, that --
`JUDGE MELVIN: Does the Kalker code then send the size for every
`block?
`MR. BURESH: Kalker sends two categories of information. It sends
`coded unit information which would be the largest smallest and intermediate
`block sizes. In other words, it establishes the values of S equals three, the
`value of S equal two and the value of S equals one. So, that's one category
`and that's the coded unit which allows the decoder to establish the
`segmentation grid and also to understand the codes that it is going to be
`receiving for this particular coded unit.
`The second category of syntax information that Kalker transmits and
`receives is the individual partitioning information for the blocks and those
`would be the 3’s, 2’s, 1’s, that attach to a particular starting block or current
`block and then indicate the partitioning of that starting size current block
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`
`
`15
`
`
`
`IPR2019-00757
`Patent 9,930,365 B2
`
`into a smaller block.
`JUDGE ULLAGADDI: Point number 2 where -- go ahead Judge
`Moore.
`JUDGE MOORE: I was just going to ask where in the decoding does
`Kalker use the size of every block?
`MR. BURESH: So, in the decoding in the process of recreating the
`segmentation map that was created at the encoder it utilizes the 3’s, 2’s that
`the streams of syntax elements that you see for example at column 5 and line
`60, the 3, 2, 2, 1, 1, 1, 2, 3, 3 in the sequence, that sets of syntax elements is
`what allows the decoder to create the segmentation map that is depicted in
`figure 9.
`So, you match up the syntax elements that indicate the particular
`partitioning of these blocks, the results is figure 9 and that figure 9 is what
`was generated by the encoder and transmitted through the two components
`of syntax information to allow the decoder to recreate that segmentation map
`that you're seeing in figure nine.
`JUDGE ULLAGADDI: Can you point us to where in Kalker it
`discloses that the mapping between S equals one or how it's transmitted or
`where it's transmitted, where in Kalker is this discussed?
`MR. BURESH: Sure. Yes. I'm going to come back to slide 23 and
`24 in direct response to your question, Your Honor. Turn to slide 25. Here
`the discussion, I think in the institution decision Your Honors asked a good
`question which is essentially how it -- when an ordinary skill in the art
`understands that the Kalker system is not hard coded such that no syntax
`information it sends or in order to establish that relationship if you will, S
`equals one equals 4 byand S equals three equals 16 by 16. So, the good
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`
`
`16
`
`
`
`IPR2019-00757
`Patent 9,930,365 B2
`
`question is and the answer is twofold. I point to and have pointed to several
`things in Kalker that indicate to one of ordinary skill in the art that Kalker
`transmits syntax information to establish that value or that relationship. The
`first and I'm going to start from kind of the least critical to the most critical.
`Least critical in Kalker is a variable walk size system. I acknowledge
`that could be interpreted in two ways. Variable walk size within a particular
`coded unit versus that the encoder and decoder can communicate between
`coded units that there are different walk sizes at play. The second and again
`just kind of working on the priority scale, there are different examples of
`block size configurations in Kalker.
`So, the example we've spoken primarily on has the by 4 is the smallest
`and the 16 by 16 is the largest,but in column five you see another example
`where the smallest S equals one, this is lines 18 through 21 of column 5.
`Where S equals ones and notes 4 by 8 blocks and S equals three denotes 8
`by 8 blocks. Again, indicative, but not conclusive.
`Claim one, however, in fact, each is an independent claim has a
`concept in it. This concept is also in the summary of the invention app at
`column one, lines 45 through 50. If you look at and I refer to slide 25
`because it's easier for me to read. But slide has claim one which is an
`inquiry of those references I just provided to you.
`In the method of transmitting encoded video pictures of claim one in
`Kalker one of the last element here is transmitting the coded picture data and
`the coded segmentation map characterized in that the step of encoding said
`segmentation map comprises one, assigning a block size code to each block
`size and in two, scanning the segmentation map in accordance with a
`predetermined scan encoder.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`
`
`17
`
`
`
`IPR2019-00757
`Patent 9,930,365 B2
`
`
`That step in there that I refer to as one of assigning a block size code
`to each block size and those block size if you look up in the first element of
`the claim you segment each picture into a map of picture blocks having
`variable block sizes. So, you have the block sizes established again based
`upon the rate distortion analysis.
`You then have this step of assigning a code. So, that's your S code to
`each block size that you just determined based upon your segmentation. So,
`that step of assigning a block size code to each block size is the opposite of a
`hardwired approach. You would not be assigning a block size code to a
`block size.
`In other words, I am seeing my segmentation into variable block sizes.
`I'm seeing that at the encoder and I'm saying I have a 16 by 16 block, I need
`to assign a block size code to it. Three is it becomes 16 by 16.
`Or I get my next picture image and the largest block size is 8 by 8.
`And that's another example from column 5. In -- that --
`JUDGE MOORE: Couldn't that just mean that for each block you're
`giving it a 1, 2 or 3 because where does it say that 1, 2 or 3 is different from
`picture to picture?
`MR. BURESH: That is the interpretation that Patent Owner presented
`and I think it's wrong for two reasons. One, it's not the words in Kalker. If
`you were assigning a block size code to each block I would agree with you,
`but that's not what's happening in the second step there. The scanning of the
`segmentation map in accordance with the predetermined scan order to obtain
`the one-dimensional series of block size code, that is where -- that is where
`you are assigning the same block size code to a particular block.
`That is different than establishing the relationship between the block
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`
`
`18
`
`
`
`IPR2019-00757
`Patent 9,930,365 B2
`
`size code and the block size. And that is what is called out in claim one and
`again also in the summary of the invention. That is establishing the
`relationship between a three and a block size, 16 by 16.
`Between a one and a block size, 4 by 4, that is not the same thing as
`establishing -- now three which I have already said to be 16 by 16 and -
`block. Those are two different concepts. And again, just the express
`language used here, and the express wording here is block sized code to
`block size. That is creating a relationship.
`JUDGE MOORE: Once again, Kalker's description that supports
`your reading of that language? And can you point us to where that is
`described?
`MR. BURESH: The -- again, just kind of walking backwards through
`what I've just gone through and again from the perspective of an ordinary --
`person of ordinary skill in the art when you are assigning a block size code
`for a block size you are doing that because you have variable block sizes in
`your various pictures.
`So, as you transition from one set of block sizes in one picture, one
`coded unit, to the next you are again you are working through assigning
`these values. If you didn't have -- if it was preordained that three equals 16
`
Accessing this document will incur an additional charge of $.
After purchase, you can access this document again without charge.
Accept $ Charge
Still Working On It
This document is taking longer than usual to download. This can happen if we need to contact the court directly to obtain the document and their servers are running slowly.
Give it another minute or two to complete, and then try the refresh button.
A few More Minutes ... Still Working
It can take up to 5 minutes for us to download a document if the court servers are running slowly.
Thank you for your continued patience.
This document could not be displayed.
We could not find this document within its docket. Please go back to the docket page and check the link. If that does not work, go back to the docket and refresh it to pull the newest information.
Your account does not support viewing this document.
You need a Paid Account to view this document. Click here to change your account type.
Your account does not support viewing this document.
Set your membership
status to view this document.
With a Docket Alarm membership, you'll
get a whole lot more, including:
- Up-to-date information for this case.
- Email alerts whenever there is an update.
- Full text search for other cases.
- Get email alerts whenever a new case matches your search.
One Moment Please
The filing “” is large (MB) and is being downloaded.
Please refresh this page in a few minutes to see if the filing has been downloaded. The filing will also be emailed to you when the download completes.
Your document is on its way!
If you do not receive the document in five minutes, contact support at support@docketalarm.com.
Sealed Document
We are unable to display this document, it may be under a court ordered seal.
If you have proper credentials to access the file, you may proceed directly to the court's system using your government issued username and password.
Access Government Site
