[00:00:32] Speaker 01:
The third case this morning is number 16-1830, NRA AT&T intellectual property 2, Mr. Trellar.

[00:00:45] Speaker 04:
Thank you, Your Honor, and may it please the Court.

[00:00:48] Speaker 04:
This inter-party's re-exam should never have started.

[00:00:52] Speaker 04:
The third party, LG, that filed the initial request for the re-exam abandoned it before the PTO acted on it.

[00:00:59] Speaker 01:
There's no question.

[00:01:03] Speaker 01:
the PTO could have maintained an ex-party re-exam based on these same facts.

[00:01:10] Speaker 01:
In other words, they could have just said, okay, well, the inter-parties re-exam is over with, but we're going to do an ex-party re-exam.

[00:01:20] Speaker 01:
They could do that, right?

[00:01:21] Speaker 04:
Perhaps, but that's not what they did.

[00:01:22] Speaker 04:
It's not perhaps.

[00:01:23] Speaker 01:
They could, right?

[00:01:24] Speaker 01:
If they made the determination that there was a reasonable

[00:01:28] Speaker 01:
to believe that there was a lack of invalidity, right?

[00:01:32] Speaker 04:
Sure, there's obviously a statutory procedure for ex parte re-exams, but that's not the procedure that was invoked or followed here.

[00:01:39] Speaker 04:
Well, what difference does it make?

[00:01:41] Speaker 04:
Well, first of all, Congress set up two separate ways of proceeding.

[00:01:46] Speaker 04:
The PTO followed one way.

[00:01:48] Speaker 04:
It didn't follow the rules for that way.

[00:01:50] Speaker 01:
Does it practically matter what difference is there?

[00:01:52] Speaker 04:
Well, there are differences between the ex parte and the inter partes reexamples.

[00:01:58] Speaker 01:
But as far as you're concerned, what difference?

[00:02:01] Speaker 01:
How are you prejudiced by the fact that they continued the inter partes proceeding rather than started a new ex parte proceeding?

[00:02:09] Speaker 04:
Well, Your Honor, we're prejudiced in a number of ways, I think.

[00:02:12] Speaker 04:
First of all, the way this started based on a piece of prior art called Yang,

[00:02:18] Speaker 04:
which was in the LG request, but which was abandoned before the PTO could act on it.

[00:02:24] Speaker 04:
Then they switched to a reading of Krauss that... They could have started a proceeding based on Krauss.

[00:02:29] Speaker 04:
They could have, but LG couldn't have pursued Krauss.

[00:02:33] Speaker 04:
So this proceeding just took off in a direction it could not have gone had it proceeded as it was supposed to.

[00:02:39] Speaker 01:
I don't think you're addressing my question.

[00:02:40] Speaker 01:
What is the difference of their continuing the inter-partis proceeding versus an ex-party proceeding?

[00:02:48] Speaker 01:
cited the Krauss prior art.

[00:02:50] Speaker 01:
They could have had a re-exam based on it.

[00:02:54] Speaker 01:
How does it adversely affect you that they did it one way rather than the other?

[00:02:59] Speaker 04:
Well, Your Honor, I guess all I can say is they invoked a particular procedure under the statute.

[00:03:06] Speaker 04:
The statute imposes requirements before they can invoke that procedure.

[00:03:10] Speaker 04:
Those requirements were not met.

[00:03:12] Speaker 04:
Perhaps they could have taken a different route.

[00:03:15] Speaker 04:
They didn't.

[00:03:17] Speaker 04:
They're supposed to comply with the statute and they didn't do it.

[00:03:21] Speaker 01:
But there are circuit cases saying that an agency can, for example, if it could, there's an EEO case, EEOC case where the agency intervened in a private proceeding even though the private party gave up.

[00:03:40] Speaker 01:
And the court said, well, the EEOC could have started independent proceedings.

[00:03:44] Speaker 01:
That doesn't make any difference that the private party gave up.

[00:03:47] Speaker 01:
Well, it seems to be similar to this.

[00:03:49] Speaker 04:
Well, it may be similar, Your Honor, but it's not the same thing, because here you have two statutorily defined separate procedures.

[00:03:59] Speaker 04:
And I don't want to beat a dead horse, but the PTO invoked one.

[00:04:08] Speaker 04:
It didn't comply with the statutory requirements for that one.

[00:04:11] Speaker 04:
And so it should not have proceeded.

[00:04:13] Speaker 04:
This proceeding should never have started.

[00:04:16] Speaker 04:
I don't want to belabor that point.

[00:04:18] Speaker 04:
Let me turn to the merits.

[00:04:23] Speaker 04:
The prior ARC and the 071 patent, the AT&T patent, both addressed the problem of processing and transmitting the huge amount of data required to send digital images.

[00:04:37] Speaker 04:
The 071 patent addresses that problem by using coding techniques that allow for data compression and by sending a single entity or vector

[00:04:45] Speaker 04:
that identifies the locations of all of the coded coefficients in a block of image data.

[00:04:53] Speaker 04:
Krauss took a very different approach.

[00:04:55] Speaker 04:
Krauss breaks up a block of image data into groups or subsets consisting of less than all of the coefficients and sends multiple vectors to then allow the block to be reconstructed.

[00:05:08] Speaker 04:
Now Krauss's preferred embodiment is to divide the block into specific regions and send a separate vector for each one

[00:05:14] Speaker 04:
But the Krauss method, whether it uses subdivided regions or not, is to send only groups or subsets of the transform coefficients at one time, rather than send them all in a single entity.

[00:05:26] Speaker 01:
But the PTO found that it disclosed doing a single block, right?

[00:05:31] Speaker 04:
Well, here's what the PTO, the PTAB found.

[00:05:37] Speaker 04:
The PTAB relied on two grounds.

[00:05:40] Speaker 04:
One was a passage at the bottom of column seven going on to column eight where Krauss says... But they made the determination, right?

[00:05:47] Speaker 01:
That Krauss disclosed doing it with respect to an entire block.

[00:05:51] Speaker 04:
Yes, they did.

[00:05:53] Speaker 04:
Sure.

[00:05:53] Speaker 04:
But they were wrong.

[00:05:55] Speaker 04:
They misread Krauss and I would submit they also applied obviousness type reasoning to an anticipation case.

[00:06:02] Speaker 04:
They focused on two things.

[00:06:04] Speaker 04:
The first was a passage where Krauss says,

[00:06:07] Speaker 04:
If you're dealing with a whole block, that's a lot of combinations of image data.

[00:06:11] Speaker 04:
And it is not easy to compute all of that.

[00:06:14] Speaker 04:
So my solution is to divide the block up into regions, and you send smaller sets of data with vectors for each.

[00:06:22] Speaker 04:
Now, the board took that not easy comment and said that reflects that it was contemplated in the prior art that you could do this.

[00:06:30] Speaker 04:
Well, first of all, saying that something is contemplated in the prior art is not a disclosure.

[00:06:36] Speaker 04:
of a specific way of doing it.

[00:06:37] Speaker 04:
What Krauss was doing was identifying the problem.

[00:06:40] Speaker 04:
The problem is, this is too much data to deal with, and my solution is, let's break it up.

[00:06:46] Speaker 04:
Moreover, even that passage doesn't disclose sending a single vector for all of the combinations or all of the non-zero and zero coefficients.

[00:06:59] Speaker 04:
It just talks about, in the abstract, the problem of computing all the combinations at one time.

[00:07:04] Speaker 04:
And then what the PTO did was it said, effectively, yeah, Kraus talks about doing this for a region and sending a vector for a region.

[00:07:15] Speaker 04:
But we think you could, essentially, you could extend this to a whole block.

[00:07:19] Speaker 04:
And it cites a passage in column 8.

[00:07:22] Speaker 04:
And the column 8 passage, it just talks about, and this is an appendix page 7 where they have this analysis.

[00:07:28] Speaker 04:
The column 8 passage just talks about dealing with a region and sending a 16-bit vector

[00:07:34] Speaker 04:
for the region, and a 16-bit vector can't transmit all the image data for an entire block.

[00:07:38] Speaker 04:
And that's all Krauss talks about throughout the specification.

[00:07:42] Speaker 04:
And in fact, the PTO in its brief, I think it's at page 19, says that Krauss discloses using a 16-bit vector to transmit the coefficients in the list of non-zero coefficients.

[00:07:59] Speaker 04:
You can't do an entire block with a 16-bit vector.

[00:08:01] Speaker 04:
That is, in fact, what Krauss discloses.

[00:08:03] Speaker 04:
But you can't do it that way.

[00:08:05] Speaker 04:
The other thing the PTO said, and this is why I say they let obviousness type considerations get into an anticipation analysis.

[00:08:15] Speaker 04:
They say, and then the PTO echoes this in their brief, there's nothing in Crouse that says you couldn't apply his method to an entire block.

[00:08:24] Speaker 04:
Well, anticipation requires an affirmative disclosure of the claimed invention, not a general statement that I don't think it's impossible.

[00:08:32] Speaker 04:
Moreover, Krauss's method is, whether he's dealing with a block or a region, is to generate a vector representing fewer than all of the non-zero transform coefficients and send them basically in units.

[00:08:48] Speaker 04:
The preferred embodiment is to divide the block into formal regions.

[00:08:53] Speaker 04:
But the concept of Krauss is breaking up the transform coefficients into sets.

[00:08:59] Speaker 04:
Subsets or groups is the term Krauss uses.

[00:09:02] Speaker 04:
and to send those rather than try to send it all at once.

[00:09:06] Speaker 04:
The other thing that the PCAB pointed to was claims one and two of Krauss.

[00:09:12] Speaker 04:
It said that claim two talks about dividing it into regions and having a vector for each.

[00:09:17] Speaker 04:
And that claim one doesn't say you have to divide it into regions.

[00:09:21] Speaker 04:
Well, that's true.

[00:09:22] Speaker 04:
But what claim one is talking about

[00:09:24] Speaker 04:
is a group of coefficients qualified for transmission from the block.

[00:09:27] Speaker 04:
Not all the coefficients qualified, not the coefficients qualified, a group of them.

[00:09:33] Speaker 04:
Krauss uses group and subset throughout the specification interchangeably to refer to some but not all of the non-zero coefficients in the block.

[00:09:42] Speaker 04:
So what Krauss is talking about is not the 0, 7, 1 invention where you code everything together and send a single vector to allow you the decoder

[00:09:53] Speaker 04:
to place the transform coefficients in the right place.

[00:09:56] Speaker 04:
It's talking about a different method, which is sending separate vectors for different subsets of the transform coefficients.

[00:10:04] Speaker 04:
Now, on appeal, the PTO, I think, recognizing the problems with the board's analysis, has an alternative argument, which they admit the board did not expressly discuss.

[00:10:15] Speaker 04:
The first problem with that, of course, is that you can only review the rationale that the board did actually discuss.

[00:10:21] Speaker 04:
So if there's any confusion about that,

[00:10:23] Speaker 04:
At a minimum, the case should be sent back for the board to explain its reasoning.

[00:10:27] Speaker 04:
In fact, the passage that the PTO relies on, which is in column 7, was not discussed at all by the board.

[00:10:35] Speaker 04:
The PCAP says, well, the board cited a part of the examiner's answer that kind of refers to it.

[00:10:43] Speaker 04:
But that's not good enough.

[00:10:44] Speaker 04:
There was no discussion of it at all.

[00:10:47] Speaker 04:
And moreover, it doesn't solve the problem.

[00:10:49] Speaker 04:
Because the passage that the PTO relies on on appeal

[00:10:53] Speaker 04:
And this is column 7, lines 27 to 47 of Krauss, appendix 49.

[00:10:59] Speaker 04:
Again, it's talking about, first of all, in large part, it's talking about the preferred embodiment, which is the subsets and regions and sending a vector for each.

[00:11:08] Speaker 04:
But even beyond that, it is talking about the concept of sending fewer than all at once with the preferred embodiment of dividing it up formally and having vectors for each.

[00:11:23] Speaker 04:
As a matter of process, it's improper because it's not the rationale the board adopted.

[00:11:27] Speaker 04:
And as a matter of substance, the PTO is wrong about what that teaches.

[00:11:34] Speaker 02:
Counsel, I'd like you to kind of put in a capsule your argument.

[00:11:39] Speaker 02:
It seems to me that the board found, and the question is whether this is substantial evidence that Krauss discloses vector coding of the entire pixel block.

[00:11:51] Speaker 02:
That's the issue that we're looking at here.

[00:11:54] Speaker 02:
Correct.

[00:11:55] Speaker 02:
And the board comments on that.

[00:11:58] Speaker 02:
We're looking at that, and we're looking at this under a standard review of substantial evidence.

[00:12:03] Speaker 04:
Correct.

[00:12:05] Speaker 02:
Point to me why that is not, why the reasoning of the board is not supported by substantial evidence.

[00:12:13] Speaker 04:
So the board's reasoning is at pages six and seven of the appendix.

[00:12:17] Speaker 04:
And what the board says is

[00:12:20] Speaker 04:
As I said, it says two things.

[00:12:22] Speaker 04:
One is it refers to that passage where Krauss says it's not easy to compute all of this.

[00:12:29] Speaker 04:
In that passage, Krauss does not say how to do it.

[00:12:36] Speaker 04:
It doesn't talk about sending a single entity or single vector for the entire block.

[00:12:40] Speaker 04:
That's the invention of the 07-1 patent.

[00:12:42] Speaker 04:
Krauss says, pardon me, this is hard to do because of the large number of potential combinations.

[00:12:48] Speaker 04:
So I've got a better way.

[00:12:48] Speaker 04:
I've got a different way.

[00:12:50] Speaker 04:
And my way is multiple regions and vectors for a single block.

[00:12:56] Speaker 04:
So that doesn't do it.

[00:12:58] Speaker 04:
The other thing the board says, let me find the exact passage here.

[00:13:08] Speaker 04:
The board says that the examiner finds that there's nothing within Krauss that suggests that the Krauss's technique must only be applied to subregions.

[00:13:15] Speaker 04:
Well, that's not an anticipation finding.

[00:13:17] Speaker 04:
It may not say, well, you can't do this.

[00:13:20] Speaker 04:
But it doesn't say you can't, which is what you'd have to have for anticipation.

[00:13:26] Speaker 04:
And finally, when the board talks about.

[00:13:30] Speaker 04:
Are you still on page six or seven?

[00:13:31] Speaker 04:
Oh, I'm sorry.

[00:13:32] Speaker 04:
I moved to page seven.

[00:13:33] Speaker 04:
I'm sorry, Your Honor.

[00:13:33] Speaker 04:
All right.

[00:13:34] Speaker 04:
But the reasoning spans those two.

[00:13:38] Speaker 04:
But there is nothing in Krauss.

[00:13:40] Speaker 04:
And the board doesn't point to anything where it's talking about a single vector for all the image data in the entire block.

[00:13:47] Speaker 04:
And saying that Krauss's method can be applied to a whole block, which the board says, well, that's true in a sense, but Krauss's method is having subsets or groups of the data, not all of it at once.

[00:14:01] Speaker 04:
So whether you say you're applying his method to the entire block or to regions within the block, it's still not the 071 patent because it's not generating a single entity for all of the image data.

[00:14:14] Speaker 00:
Thank you, your honor.

[00:14:19] Speaker 00:
Mr. Foreman.

[00:14:28] Speaker 03:
May it please the court.

[00:14:29] Speaker 03:
Substantial evidence supports the anticipation rejection in this case.

[00:14:34] Speaker 03:
If you look at the Kroos prior patent, specifically appendix page 49, column 7, line

[00:14:43] Speaker 03:
27 says, in accordance with the present invention, a unique code word is used to specify a subset of coefficients selected for transmission within a block of transformed coefficients or within portions of the block referred to as regions.

[00:14:58] Speaker 03:
Later on in that same column, there are two other references to block or region.

[00:15:03] Speaker 03:
And then at line 51, it explains that only coefficients with amplitudes greater than 0

[00:15:12] Speaker 03:
after quantization are transmitted.

[00:15:14] Speaker 03:
So this is a disclosure right here of Krauss saying that the vector coding method can be applied to a block of transform coefficients or a region of transform coefficients.

[00:15:28] Speaker 03:
And while much of Krauss is discussing the application to the regions.

[00:15:36] Speaker 02:
Let's go back to around line 28.

[00:15:40] Speaker 02:
Sure.

[00:15:41] Speaker 02:
In accordance with the present invention, a unique code word is used to specify, yeah, it goes on the coefficients selected within a block.

[00:15:52] Speaker 02:
And you're saying that the use of the block here signifies single coding?

[00:16:00] Speaker 03:
Yes.

[00:16:01] Speaker 03:
It's saying this can be applied to a block of transform coefficients, or it can be applied to regions.

[00:16:08] Speaker 02:
But it says within portions of the block.

[00:16:11] Speaker 02:
And it describes those portions as regions.

[00:16:14] Speaker 03:
Correct.

[00:16:16] Speaker 03:
The regions part is what AT&T is focusing on.

[00:16:20] Speaker 03:
And that, we acknowledge, is not what AT&T's claim one is about.

[00:16:26] Speaker 03:
And CROS spends much of its time discussing this regions embodiment, where it takes a square block and breaks it up into.

[00:16:34] Speaker 01:
You're saying that that sentence is distinguishing within a block, meaning the whole block, or within a portion

[00:16:41] Speaker 03:
Yes, it's saying it can be applied to the full block, or it can be applied to regions within the block.

[00:16:46] Speaker 03:
There are two different options there.

[00:16:47] Speaker 03:
And then later on, it says line 38 says block a region.

[00:16:52] Speaker 03:
Line 44 says each block a region.

[00:16:55] Speaker 03:
And this is consistent with Claim 1, Claims 1 and 2 of Crouse.

[00:16:59] Speaker 02:
I'm sorry, I missed it.

[00:17:01] Speaker 02:
Show me where it says the whole block.

[00:17:05] Speaker 03:
Well, starting.

[00:17:09] Speaker 03:
Lines 29 and 30 selected for transmission within a block of transform coefficients or within portions of the block referred to as regions.

[00:17:18] Speaker 03:
So what it's doing is it's looking for the non-zero transform coefficients.

[00:17:24] Speaker 03:
So it's either looking for them within a full block or it's looking for them within a specific region within the block.

[00:17:30] Speaker 03:
And then later on, line 38 says the group of coefficients that could be selected in a block or region.

[00:17:38] Speaker 03:
And then line 44 says,

[00:17:40] Speaker 03:
the group of coefficients selected for transmission from each block or region.

[00:17:44] Speaker 03:
So it's clearly referring to two different things.

[00:17:46] Speaker 03:
You can apply this to a block, or you can apply this to a region within a block.

[00:17:50] Speaker 03:
And claims one and two echo that, because claim one states that you provide a block of transformed coefficients and generate a vector to identify the non-zero coefficients.

[00:18:01] Speaker 02:
What about the statement at 25 that says,

[00:18:06] Speaker 02:
there's no particular scanning method is required to maintain the high compression.

[00:18:11] Speaker 03:
That's consistent with AT&T's patent, which does the same thing.

[00:18:16] Speaker 03:
But that's not part of, I mean, that's not recited in AT&T's claim.

[00:18:21] Speaker 03:
So I think what AT&T seems to be focusing on is the subset language.

[00:18:28] Speaker 03:
But I think all that refers to is that Krauss's method could be applied to

[00:18:36] Speaker 03:
identify any sort of transform coefficients that are selected for transmission.

[00:18:42] Speaker 03:
It gives the example starting at line 51 that that could be non-zero coefficients.

[00:18:47] Speaker 03:
But they just describe it more generally that it could be any subset that you choose.

[00:18:53] Speaker 03:
So both in AT&T's patent and here, one of the example subsets is all non-zero coefficients.

[00:19:01] Speaker 03:
So it's doing the same thing.

[00:19:02] Speaker 03:
It's looking at a block of transform coefficients

[00:19:05] Speaker 03:
finding the non-zero coefficients and creating a vector to identify those locations.

[00:19:12] Speaker 03:
And this is stated, again, in column 7.

[00:19:16] Speaker 03:
This is stated with the example of the 8 by 8 block that the examiner and the board relied on.

[00:19:22] Speaker 03:
And this is stated within claims 1 and 2 of PROS.

[00:19:26] Speaker 03:
And we think that's more than enough to meet the substantial evidence requirement.

[00:19:32] Speaker 03:
Now, I acknowledge, and I acknowledge this in my brief,

[00:19:35] Speaker 03:
that the board doesn't specifically talk about the blocker region language in column 7 of Crouse.

[00:19:46] Speaker 03:
But the board approvingly cites to the examiner's answer, which discusses this in detail.

[00:19:52] Speaker 03:
And this column 7 language was always part of the examiner's rejection.

[00:19:57] Speaker 03:
And there's nothing in the board's decision that goes against that.

[00:20:00] Speaker 03:
And the board's decision is fully consistent with that.

[00:20:02] Speaker 03:
So we don't think this is a situation where

[00:20:07] Speaker 03:
going beyond what the board's decision is, or saying something that's not in the board decision.

[00:20:12] Speaker 03:
I mean, it's not expressly in there, but the board cites the examiner's answer, the examiner expressly relies on it, and the examiner had relied on it both in the action closing prosecution, the right of appeal notice, and the examiner's answer.

[00:20:29] Speaker 03:
Now, if there are no further questions on the anticipation or rejection, I'll just say,

[00:20:34] Speaker 03:
couple words about AT&T's procedural arguments.

[00:20:38] Speaker 03:
AT&T is trying to argue that because LG unsuccessfully tried to ask the PTO to deny its own re-exam request, that there was no active requester.

[00:20:51] Speaker 03:
But there's no mechanism within the statutes or the PTO rules for a requester to withdraw a re-exam request.

[00:21:00] Speaker 03:
So as far as

[00:21:04] Speaker 01:
What difference does it make?

[00:21:05] Speaker 01:
The PTL could have instituted an ex parte procedure.

[00:21:08] Speaker 01:
I agree.

[00:21:09] Speaker 03:
Under section 303, the director has the right to institute an ex parte re-exam on her own accord.

[00:21:17] Speaker 03:
So under that rationale, the director could have instituted a re-exam on Krauss.

[00:21:25] Speaker 02:
But that didn't happen here.

[00:21:26] Speaker 03:
There wasn't a re-institution.

[00:21:28] Speaker 03:
No, because it wasn't necessary to do that, because we weren't doing anything

[00:21:32] Speaker 03:
contrary to the statute.

[00:21:34] Speaker 03:
There was a requester here.

[00:21:36] Speaker 03:
LG filed a proper request.

[00:21:38] Speaker 03:
There was no mechanism to withdraw it or have it denied on LG's own request.

[00:21:44] Speaker 03:
And the PTO considered it and granted the re-exam request.

[00:21:49] Speaker 03:
And there was nothing, we did nothing contrary to the statute in doing that.

[00:21:55] Speaker 03:
This is not like district court litigation or even an IPR where there's some right of the

[00:22:02] Speaker 03:
the party bringing the action to withdraw it or seek that it's dismissed, once you file a re-examine request, you can't unring that bell.

[00:22:15] Speaker 03:
If you give us that re-examine request, we're going to make a determination on it.

[00:22:21] Speaker 03:
So because that re-examine request was filed and there was no mechanism for LG to withdraw it, we rendered a decision on it.

[00:22:29] Speaker 02:
Do we have jurisdiction to review this issue?

[00:22:33] Speaker 03:
We don't believe under 312C, we don't believe that, because this goes to the institution of the re-exam, we don't believe there's jurisdiction.

[00:22:46] Speaker 03:
But even if you want to reach the ultimate issue, I don't think there's any issue of us going beyond our statutory authority.

[00:22:58] Speaker 03:
There's nothing further.

[00:23:01] Speaker 03:
Thank you.

[00:23:05] Speaker 04:
Thank you, Your Honor.

[00:23:08] Speaker 04:
Very briefly, what you just heard from counsel on the merits was this new argument that they came up with on appeal.

[00:23:17] Speaker 04:
Well, wait, but it's in the examiner's answer, right?

[00:23:20] Speaker 04:
I think if you look at the examiner's answer, first of all, the board says, oh, and the examiner said that there's no reason you can't apply this method to an entire block, has a general citation to the examiner's answer.

[00:23:33] Speaker 04:
The examiner's answer,

[00:23:34] Speaker 04:
talks about column seven in very general terms doesn't focus on this language in particular.

[00:23:39] Speaker 04:
So I don't think you can say that this is, that this argument you just heard reflects the board's reasoning.

[00:23:44] Speaker 04:
It's not what the board said.

[00:23:45] Speaker 04:
The board focused on different parts of Kraus and not this part.

[00:23:50] Speaker 04:
Moreover, saying that you, as I said when I was up here before, saying that you can apply Kraus's method to a block or a region doesn't answer the question because even in that column seven passage the council was referring to,

[00:24:04] Speaker 04:
What it talks about is a subset of coefficients selected for transmission from a block or a region.

[00:24:09] Speaker 04:
Not the coefficient selected, not even the subset selected.

[00:24:13] Speaker 04:
It's a subset, and that's the key.

[00:24:15] Speaker 04:
Because what Krauss is talking about is dealing with fewer than all at one time, rather than trying to process and transmit the entire block at once.

[00:24:24] Speaker 04:
And I would direct the court's attention to column 11, the passage beginning at about line 12 through 35 or so.

[00:24:32] Speaker 04:
And I'm not going to read it to the court,

[00:24:34] Speaker 04:
But this is sort of the summary.

[00:24:36] Speaker 04:
Krause's wrap up before getting into claims.

[00:24:39] Speaker 04:
It should now be appreciated that here's my present invention.

[00:24:42] Speaker 04:
And the important part is it talks about sending vector code words to reconstruct a single block of image data.

[00:24:49] Speaker 04:
So it's talking about multiple vectors, not a single vector.

[00:24:53] Speaker 04:
And that's the nub of the invention.

[00:24:56] Speaker 01:
OK, I think we're out of time.

[00:24:57] Speaker 04:
I am.

[00:24:58] Speaker 04:
Thank you, Your Honor.

[00:24:59] Speaker 01:
I thank both counsels.

[00:25:00] Speaker 01:
The case is submitted.

[00:25:01] Speaker 01:
That concludes our session for today.

[00:25:04] Speaker ?:
ice.