[00:00:00] Speaker 02:
Opera face is Divix, if that's how you pronounce it.

[00:00:04] Speaker 02:
Versus United happens, 2023, 1699.

[00:00:10] Speaker 02:
Mr. Hendefar.

[00:00:14] Speaker 04:
Morning, Your Honors.

[00:00:15] Speaker 04:
May I please the court?

[00:00:17] Speaker 04:
I would like to start by an analogy, if I may.

[00:00:21] Speaker 04:
Imagine a claim that recites a car that is half red and half blue.

[00:00:28] Speaker 04:
Petitioners do not find a reference that divides the car into two colors, let alone the color's claimed, let alone dividing in the way the claim does.

[00:00:38] Speaker 04:
All they have is a car that is red in its entirety,

[00:00:42] Speaker 04:
and a car that is blue in its entirety.

[00:00:45] Speaker 04:
And what they argue is that a person of ordinary skill would be motivated to combine the red car and the blue car to come up with a car that is half red and half blue.

[00:00:58] Speaker 04:
But their motivation is that blue is better than red.

[00:01:02] Speaker 02:
But those aren't our facts.

[00:01:04] Speaker 04:
Correct.

[00:01:04] Speaker 04:
So our facts are a claim that divides the playback period into two phases.

[00:01:11] Speaker 04:
It uses one stream selection for phase one, and it uses a second stream selection method for phase two.

[00:01:20] Speaker 04:
Petitioners, again, do not have a reference that divides the playback method period into the two phases that the claim does.

[00:01:28] Speaker 04:
They do not have a reference that uses two different stream selection methods during playback.

[00:01:34] Speaker 04:
All they have is one reference that uses one stream selection method for the entirety of playback.

[00:01:41] Speaker 04:
and a second reference that uses a second stream selection method for the entirety of its paper.

[00:01:47] Speaker 02:
But the issue was obviousness, not anticipation.

[00:01:49] Speaker 04:
Correct.

[00:01:51] Speaker 04:
Their motivation as to why a positor would combine these two references to arrive at a claim is that the second method is better than the first method.

[00:02:01] Speaker 03:
Yeah, but I don't see Gigliotti as being that separate a method.

[00:02:09] Speaker 03:
from what's shown in Bitterman.

[00:02:11] Speaker 03:
And Gigliotti recognizes that there are lots of different ways to do it, particularly in paragraph 13 on page 1034.

[00:02:22] Speaker 03:
And it does certainly start the streaming without reference to the filling of the buffer.

[00:02:31] Speaker 03:
It starts the streaming without filling the buffer.

[00:02:36] Speaker 04:
That's respectfully not entirely accurate.

[00:02:40] Speaker 04:
If I may please refer your honor to appendix page 1026.

[00:02:44] Speaker 03:
On 1024 it should start at first connection rate.

[00:02:52] Speaker 03:
That's before

[00:02:54] Speaker 03:
checking anything about the buffer.

[00:02:56] Speaker 04:
But there is no playback yet.

[00:02:58] Speaker 03:
All I'm saying is they start in the same way that Bitterman starts, and it recognizes in paragraph 13 that there are different ways of doing this, including

[00:03:11] Speaker 03:
measuring the stream and the data rate.

[00:03:17] Speaker 03:
It isn't simply a single defined method.

[00:03:22] Speaker 03:
Galeotti offers a lot of flexibility.

[00:03:26] Speaker 04:
Right.

[00:03:26] Speaker 04:
And it is a defined method as shown in Figure 3, which is what the board relied on, which is what the petition relied on.

[00:03:33] Speaker 04:
And this is what Equiality does.

[00:03:35] Speaker 04:
It fills a buffer to a first threshold.

[00:03:38] Speaker 04:
Then it starts playback.

[00:03:41] Speaker 04:
After it starts playback,

[00:03:42] Speaker 04:
it uses the bandwidth and the buffer to decide whether it's going to switch to a different string.

[00:03:48] Speaker 04:
The issue is whether after a gigliotti starts the playback, which on figure 2, which is what you are pointed to, is step 218, the issue is whether after a gigliotti starts the playback,

[00:04:01] Speaker 04:
does it use one stream selection method or two different stream selection methods?

[00:04:06] Speaker 04:
And it has one stream selection method as shown by the algorithm.

[00:04:10] Speaker 04:
But I want to just step back for a moment.

[00:04:12] Speaker 01:
Before we move away from gigliotti, when I look at figure three and then read the discussion at 10.36 to 30 seconds, it's my understanding that gigliotti is not

[00:04:29] Speaker 01:
using its check the buffer fill level feature from the very, very beginning of playback in Gigliotti.

[00:04:38] Speaker 01:
It appears that Gigliotti has two buffers.

[00:04:41] Speaker 01:
There's an initial buffer, and then there's a larger buffer.

[00:04:44] Speaker 01:
And the initial period, it's not thinking about the larger buffer.

[00:04:50] Speaker 01:
It's just using the initial buffer for maybe two to five seconds.

[00:04:54] Speaker 01:
And then from after that period, it starts

[00:04:59] Speaker 01:
playing back from the larger buffer.

[00:05:01] Speaker 01:
And it's only with the larger buffer is Gigliotti checking for possible buffer underfill.

[00:05:08] Speaker 01:
And then if it falls below some watermark, then will it switch to a different video stream with a lower bitrate.

[00:05:17] Speaker 01:
All that to say is it doesn't appear to me that Gigliotti itself is contemplating the sensitivity about buffer underfill from the very beginning of the playback.

[00:05:27] Speaker 01:
Let me ask.

[00:05:28] Speaker 01:
any combination of Bigliotti with Bitterman wouldn't actually be necessarily checking for buffer underfill from the very very beginning.

[00:05:38] Speaker 04:
And I understand what I believe the issue may be that I need to address.

[00:05:43] Speaker 04:
The portions of the claim invention we are discussing and the portions that are at dispute here are after the playback has already started.

[00:05:51] Speaker 04:
I agree with your honor that Gigliotti has an initial buffer before it starts playback, that it fills that initial buffer, then it starts playback, and at that point, it considers the buffer.

[00:06:03] Speaker 04:
The claim limitation we are looking at

[00:06:05] Speaker 04:
are both after the playback has already started.

[00:06:08] Speaker 04:
So the question is, after the playback has already started... I'm a little lost.

[00:06:14] Speaker 01:
The claim talks about a startup period, right?

[00:06:20] Speaker 01:
So that's literally from the very beginning.

[00:06:24] Speaker 01:
It's not after the playback has begun.

[00:06:27] Speaker 04:
So within that initial startup period, there is a playback and there is a switching of districts.

[00:06:32] Speaker 04:
So if I may please refer you to appendix page 65.

[00:06:36] Speaker 04:
And that's the claim one in the patent.

[00:06:41] Speaker 04:
And column 15, line five is when the initial startup period starts.

[00:06:47] Speaker 04:
Within that initial startup period, you first select a stream, then you start playback, and then you switch that stream based on the bandwidth, all within that initial startup period.

[00:06:59] Speaker 04:
So the initial startup period includes switching the stream based on the bandwidth after the playback has already started.

[00:07:07] Speaker 04:
And that's online.

[00:07:10] Speaker 04:
18 of column 15, playing back at least one chunk of the first alternative stream stored in the buffer, obtaining- Yeah, but the initial startup period includes the selection of the first video stream, right?

[00:07:24] Speaker 01:
Correct, so it has- Storing that first video stream, or at least a chunk of it in the buffer.

[00:07:29] Speaker 01:
Yes.

[00:07:29] Speaker 01:
And then playing that back.

[00:07:32] Speaker 04:
And then switching the stream.

[00:07:34] Speaker 01:
That's 304, 306 of figure three

[00:07:38] Speaker 01:
Gagliotti, right?

[00:07:40] Speaker 04:
Sure, fine.

[00:07:41] Speaker 04:
And then the initial startup period, however, has more steps in the claims.

[00:07:45] Speaker 04:
It has additional measurement and an additional switching of the streams based on the bandwidth during the initial startup period.

[00:07:53] Speaker 01:
I guess the point still remains, though, that even if you do a straight-up combination of Gagliotti with Binomy, that wouldn't necessarily require checking buffer or underfill.

[00:08:06] Speaker 01:
levels at the very beginning of the start-up of the playback.

[00:08:12] Speaker 01:
Is that right?

[00:08:13] Speaker 04:
Well, that's not right, because Gigliotti says that you do.

[00:08:16] Speaker 01:
So if you're on a... Gigliotti doesn't start thinking about that until step 3.14, which is after its initial period ends with the initial buffer, and then you're

[00:08:26] Speaker 01:
working on the larger buffer.

[00:08:28] Speaker 04:
Right.

[00:08:28] Speaker 04:
But in the initial buffer, you do not have a playback.

[00:08:32] Speaker 04:
So the playback in nucleotide starts after the initial buffer is already filled.

[00:08:36] Speaker 04:
And that's clear on figure 2, where your honor will see step 204.

[00:08:40] Speaker 04:
That's appendix base 1026.

[00:08:42] Speaker 01:
I'm looking at step 306.

[00:08:45] Speaker 04:
Sure.

[00:08:47] Speaker 01:
Step 306.

[00:08:48] Speaker 01:
Sure.

[00:08:50] Speaker 04:
Correct.

[00:08:52] Speaker 04:
So in step 306, you start the playback, and Gigliotti uses... So now we're playing the video.

[00:09:00] Speaker 04:
Correct.

[00:09:01] Speaker 04:
Now, Your Honor is saying that after the playback, Gigliotti does not consider the buffer for a certain period of time?

[00:09:08] Speaker 01:
Is that...?

[00:09:10] Speaker 01:
I'm saying your argument is this combination of these two references would necessarily lead to

[00:09:19] Speaker 01:
a result where you must be checking for a buffer underfill from the very beginning of the startup.

[00:09:26] Speaker 04:
That's not what I'm saying.

[00:09:27] Speaker 04:
All I'm saying is, petitioners say, we divide the claimed invention into two phases.

[00:09:33] Speaker 04:
We are only going to use Bidema for phase one.

[00:09:36] Speaker 04:
And we are only going to use Gigliotti for phase two.

[00:09:39] Speaker 04:
And what we are saying is there is no motivation to use Gigliotti only for phase two of the claim.

[00:09:45] Speaker 04:
Petitioner's motivation is Gigliotti is better than Bitterman.

[00:09:49] Speaker 04:
So if that's the case, you would use all of Gigliotti for all of the time duration.

[00:09:53] Speaker 01:
If we're already working off the premise that there's substantial evidence to support the board's finding to add Gigliotti to Bitterman's playback process,

[00:10:21] Speaker 01:
it will be an improvement.

[00:10:23] Speaker 04:
But why would you choose that specific point in the meter man to add?

[00:10:26] Speaker 01:
Because the point is, almost any point in the process would improve the resulting process from that point of the addition to the end of the playback.

[00:10:38] Speaker 04:
But the lock is not that you can use the claim as a road map, add gigliotti to beater man just based on where the claim requires and then say because that's an improvement there is motivation.

[00:10:50] Speaker 04:
The test is the opposite, not doing the claim.

[00:10:53] Speaker 04:
Would the positor have been motivated to add gigliotti to beater man

[00:10:56] Speaker 04:
at that random midpoint that petitioners do, not based on anything in the references, but based on the claim itself.

[00:11:03] Speaker 04:
So I understand, Your Honor, that you can add gigliotti to any point in Biedermann, and that may be an improvement.

[00:11:08] Speaker 04:
But the issue is, would the Posita have known to add gigliotti to that specific point in the middle of playback in Biedermann?

[00:11:16] Speaker 04:
And that's the issue.

[00:11:17] Speaker 04:
There is no evidence in the record.

[00:11:18] Speaker 04:
And I would like to reiterate the board's finding.

[00:11:21] Speaker 04:
That's on Appendix 40 and 41.

[00:11:23] Speaker 04:
All the board found was that a positor would have known Gigliotti's method is superior to Biederman because it considers the buffer.

[00:11:31] Speaker 04:
That does not teach you to add Gigliotti at some random point in the middle of playback in Biederman.

[00:11:38] Speaker 04:
Even if

[00:11:39] Speaker 04:
In retrospect, that may be an improvement.

[00:11:41] Speaker 04:
The issue is not knowing the claim.

[00:11:44] Speaker 04:
A positor wouldn't have known just based on Biderman and Gigliotti.

[00:11:48] Speaker 01:
Do you agree that Gigliotti doesn't check the buffer fill level from the very beginning of the Gigliotti process?

[00:11:57] Speaker 04:
The Gigliotti process, if your honor is pointing to before the payback starts, I actually don't think I agree with that because it needs to check the fill level for it to make sure the buffer is filled.

[00:12:06] Speaker 01:
I guess my point is there's two buffers.

[00:12:10] Speaker 01:
There's an initial buffer and there's a larger buffer.

[00:12:13] Speaker 01:
It initially starts playing back from the initial buffer.

[00:12:17] Speaker 04:
Once it's filled, correct.

[00:12:18] Speaker 01:
And then it's playing from the larger buffer.

[00:12:21] Speaker 04:
Correct.

[00:12:22] Speaker 01:
And then when it's playing from the larger buffer, it's checking for buffer under fill.

[00:12:30] Speaker 01:
And it's not ever checking for buffer underfill in the initial buffer.

[00:12:34] Speaker 04:
Because it wasn't playing back when it only had the initial buffer.

[00:12:38] Speaker 04:
So the playback starts after the initial buffer is filled.

[00:12:41] Speaker 04:
So that's naturally where you check for the buffer underfill.

[00:12:45] Speaker 04:
But again, I want to point out, they're not relying on just Gigliotti and say Gigliotti's initial buffer is the first phase of the claim.

[00:12:51] Speaker 04:
Second buffer is the second phase of the claim.

[00:12:54] Speaker 04:
They are saying we are going to bring Gigliotti in the middle of Biedermann's payback and we're just going to apply it here.

[00:13:01] Speaker 04:
And how did they come up with this point in Biedermann?

[00:13:03] Speaker 04:
Based on the claim.

[00:13:04] Speaker 04:
Not based on what Biedermann says, start switching the method here.

[00:13:08] Speaker 04:
Not based on what Gigliotti says, combine us with Biedermann at this point.

[00:13:13] Speaker 04:
The point that they find in Biedermann to bring in Gigliotti is based on the claim.

[00:13:18] Speaker 04:
It's not based on our independence.

[00:13:19] Speaker 02:
I've not told you well until your rebuttal time.

[00:13:21] Speaker 02:
I assume you wish to save it.

[00:13:23] Speaker 04:
I do.

[00:13:23] Speaker 04:
And I appreciate it.

[00:13:26] Speaker 02:
Here's all of it.

[00:13:26] Speaker 04:
Thank you.

[00:13:52] Speaker 00:
Good morning and may it please the court.

[00:13:54] Speaker 00:
Angela Oliver on behalf of Unified Patents.

[00:13:58] Speaker 00:
DIV-X has not disputed that adding Gigliotti's buffer fill consideration technique to vitamin's very similar system would have been beneficial and would have improved vitamin system.

[00:14:10] Speaker 00:
Those are undisputed factual findings that we believe are sufficient to affirm the board's motivation decision in this case.

[00:14:16] Speaker 00:
Specifically, the board found that this would be beneficial because without some method,

[00:14:21] Speaker 00:
for measuring the fill level in Vitamins buffer, which it already uses, Vitamins might switch streams without having enough content in the buffer to prevent underflow.

[00:14:30] Speaker 00:
And of course, that results in interruptions to the user.

[00:14:33] Speaker 00:
The board made these findings at appendix pages 40 and 41.

[00:14:36] Speaker 00:
The board said this was a specific benefit with a rational underpinning for combining Vitamins and Gigliotti because this would avoid interruptions in streaming and thereby improve Vitamins system.

[00:14:49] Speaker 00:
So it's our contention that adding Gigliotti to vitamin system is an improvement, including, as your honor noted, at any stage of that process, it's going to be a beneficial improvement.

[00:14:59] Speaker 00:
So even if it's part of the way through, that's still a benefit, because fewer interruptions to the user are always going to be better.

[00:15:07] Speaker 01:
Did you say that part in either your petition or your reply?

[00:15:12] Speaker 01:
The point about it would be beneficial to add a Gigliotti scheme

[00:15:21] Speaker 01:
process of playback in Vinnerman?

[00:15:24] Speaker 00:
Not expressly, because we argued overall that this would be an overall benefit, adding this technique.

[00:15:29] Speaker 01:
But there's a timeline feature to this claim, where during time one, one scheme happens.

[00:15:35] Speaker 01:
During time two, another scheme happens.

[00:15:38] Speaker 01:
And so then the question is, why would a skilled artist back in 2011, with Vinnerman and Giggler

[00:15:52] Speaker 01:
Sure.

[00:15:54] Speaker 00:
And we explained in our petition that this is based on a predetermined threshold being met in Gigliotti.

[00:16:02] Speaker 00:
We referred to that initial buffer as an example of that threshold.

[00:16:06] Speaker 00:
And that would be the point in time where it makes sense to start considering the buffer fill.

[00:16:11] Speaker 00:
Because prior to that, as Your Honor noted, in Gigliotti's Figure 3, for example,

[00:16:19] Speaker 00:
Before you start filling the buffer, there's not a buffer fill level to consider.

[00:16:24] Speaker 00:
So we did talk about that being a predetermined threshold, that at that point is when you switch and you start considering the buffer fill level.

[00:16:31] Speaker 00:
Now, just to take a step back, I'm sorry, was there a question?

[00:16:36] Speaker 00:
OK.

[00:16:36] Speaker 00:
Just to take a step back, we consistently argued through our briefing and here on appeal that this is simply not the relevant question for the obviousness analysis.

[00:16:45] Speaker 00:
There's no need to show why you would not also consider the buffer fill level during phase one.

[00:16:51] Speaker 00:
And that's simply not something required by the claims.

[00:16:54] Speaker 00:
It's not something required under this court's case law.

[00:16:57] Speaker 00:
So for example, obviousness does not require a party to address other potential changes that might have been made when implementing the combination.

[00:17:05] Speaker 00:
We've discussed the GE versus Raytheon for that principle.

[00:17:08] Speaker 00:
Obviousness also does not require showing the best combination, as long as it's a suitable combination.

[00:17:15] Speaker 00:
So even if considering the buffer fill level during the phase one switch would have been better, would have been also beneficial, that's simply not something that needed to be shown here to establish obviousness based on the claim language and based on the court's case law.

[00:17:35] Speaker 00:
And in particular, we'll also note just that this was addressed as a matter of motivation throughout the proceedings below.

[00:17:45] Speaker 00:
But again, it's our position that this was not relevant to the obvious analysis because there was no requirement in the claims or in the obviousness case law to establish this separate point about why buffer fill level should not be considered.

[00:17:58] Speaker 00:
The board said at appendix page 41 that what they're really asking is for a negative motivation to show why you would not do

[00:18:05] Speaker 00:
this beneficial method during the first period.

[00:18:11] Speaker 00:
And the last thing I'll note, Your Honor, is with respect to how the 987 patent describes this initial startup period, which of course is how we read the claims, there is a discussion, a brief discussion of the initial startup period and the specification, which does consider the buffer fill level during that process.

[00:18:33] Speaker 00:
That's in column 13.

[00:18:35] Speaker 00:
Line 36, if that's useful for the court.

[00:18:37] Speaker 00:
Line 36 through 41.

[00:18:40] Speaker 00:
Again, it's our position that the specific benefit that the board found in this case of adding Gigliotti's buffer fill level consideration to the technique in bitumen would have been specifically beneficial, and that is sufficient to affirm obviousness here.

[00:18:56] Speaker 00:
Unless.

[00:18:57] Speaker 01:
Is this patent in any litigation?

[00:19:01] Speaker 00:
Not at this time.

[00:19:03] Speaker 00:
I don't know particularly the strategic cause of why.

[00:19:09] Speaker 00:
And at this point in time, this case is also quite old.

[00:19:13] Speaker 00:
This was filed many, many years ago.

[00:19:16] Speaker 00:
But the patent is not currently in litigation.

[00:19:19] Speaker 00:
And I'll say there are only two claims at issue here, claims one and 10.

[00:19:24] Speaker 00:
There are many dependent claims that were not challenged.

[00:19:27] Speaker 00:
Many of those claims address what

[00:19:30] Speaker 00:
With the more detailed aspects that the patent discusses regarding the upper bound seek delay and the lower bound seek delay, for example, those are in claims three and five.

[00:19:39] Speaker 00:
But those concepts are just not at issue in this proceeding.

[00:19:43] Speaker 02:
The patent isn't about to expire.

[00:19:46] Speaker 00:
I don't believe so.

[00:19:48] Speaker 02:
There's a fair amount to go.

[00:19:52] Speaker 02:
Anything further?

[00:19:54] Speaker 00:
Not for me, Your Honor.

[00:19:55] Speaker 00:
Thank you.

[00:19:56] Speaker 00:
Thank you.

[00:20:05] Speaker 02:
Mr. Hanifar, we'll give you three minutes for a bottle if you need one.

[00:20:09] Speaker 04:
I appreciate that very much.

[00:20:11] Speaker 04:
So Your Honor asked my friend on the other side, did Unified explain a motivation as to why a posita would bring gigliotti at this random point in Peter Man's playback, and is there a motivation for that?

[00:20:24] Speaker 04:
And counsels did not expressly.

[00:20:25] Speaker 04:
Let me read for you what petitioner argued before the board.

[00:20:28] Speaker 04:
That appendix page number 1754 and 1755, their motivation is that,

[00:20:35] Speaker 04:
A Posita would have recognized that without a technique for measuring the amount of content in the buffer, Beaterman's client might switch to the second playlist content without sufficient content stored in the buffer, which might result in interruption.

[00:20:50] Speaker 04:
That's not a motivation to bring in gigliotti at some random point in Biderman's payback, absent the claim's teaching.

[00:20:58] Speaker 04:
That's a teaching for gigliotti being a wholesome substitute for Biderman, exactly what gigliotti itself does.

[00:21:06] Speaker 04:
So there is no evidence in the record for the board to find that we are bringing in gigliotti, but not for the entirety of the claim, just for this portion of the claim.

[00:21:15] Speaker 04:
The only reason they do that is to use the claim as a roadmap.

[00:21:20] Speaker 04:
Now, I also want to highlight the significance of using the two different stream selection methods on switching between them in the patent.

[00:21:28] Speaker 04:
And the embodiment that's relevant here is the embodiment of figure five, which starts on column 12, line 47, appendix page 63.

[00:21:37] Speaker 04:
And the patent explains that during an initial startup period, it uses one stream selection method, starts the playback,

[00:21:44] Speaker 04:
then it uses that same stream selection method at an initial stage of playback.

[00:21:49] Speaker 04:
And then at some point later, it switches the stream selection method to a different method.

[00:21:54] Speaker 04:
So two different stream selection methods.

[00:21:56] Speaker 04:
And that's also shown in the prosecution of the pattern.

[00:22:03] Speaker 04:
where the applicant explained to the examiner the significance of using the two stream selection methods.

[00:22:10] Speaker 04:
So the issue here is, why would a prositer, absent the claim, would have brought in gigliotti, not to use gigliotti as it itself does, not to use vitamin as it itself does, but know, the prositer would know without the claim,

[00:22:24] Speaker 04:
to use two different stream selection methods for the two different playback periods.

[00:22:29] Speaker 04:
How would a visitor have known to bring in Gigliotti at this point?

[00:22:32] Speaker 02:
You keep saying proceeded.

[00:22:33] Speaker 02:
To speak English in court, I think you're talking about a person of ordinary skill.

[00:22:37] Speaker 04:
Yes, I apologize.

[00:22:38] Speaker 04:
How would that person of ordinary skill in the art, not knowing the claim, have known to take Gigliotti, not use Gigliotti's own algorithm, but bring it in Biedermann, not for the entirety of the playback,

[00:22:51] Speaker 04:
but some random point in the middle of the playback just based on what the client says.

[00:22:55] Speaker 04:
There is no teaching in the record.

[00:22:57] Speaker 04:
There is no motivation argued by petitioner.

[00:23:00] Speaker 04:
And the board's motivation that the board found did not provide a motivation for some middle point in playback.

[00:23:06] Speaker 04:
It would be a motivation throughout the playback.

[00:23:09] Speaker 04:
And with that, if there are any questions, I'd be happy to answer.

[00:23:12] Speaker 02:
Thank you, counsel.

[00:23:14] Speaker 02:
To both counsel, the case is submitted.