Guest Post: Regnoicorp -- A Miscarriage of Justice Calling for En Banc Reconsideration

By Paul Cole* --

Study of the background to Recognicorp, LLC v. Nintendo Co. (Fed. Cir. 2017), the subject of Michael Borella's earlier posting, reveals basis for his concerns about lack of analysis of the detailed disclosure of the specification and inattention to the detailed subject matter which was actually claimed.

A reader of Judge Reyna's opinion is prompted to question why a panel including Judges Lourie and Stoll, both experienced in patent matters, treated the invention dismissively.  Recognicorp LLC which is the assignee of the '303 patent is an entity set up to undertake patent enforcement.  A web search reveals nothing relevant about that company or its activities.  The panel may therefore have been misled into believing that Recognicorp was a mere troll, and that there was therefore nothing significant that could be said in favour of the patent and the entity owning it.

Further investigation reveals that this is an incomplete picture.  The earlier owner of the '303 patent is IQ Biometrix, Inc. of Kingwood, Texas whose president Greg J. Micek has said that the inspiration for setting up the company was the 9/11 events, and has explained that the company retains at least a contractual interest in the '303 patent.  It provides law enforcement and security technology solutions for government and private industry and its product FACES allows a user to create and re-create billions of human faces.  Facial features selected from a database are automatically blended together to produce a photo-quality composite facial image as is apparent from the images, see their website.  The technology helps law enforcement agencies identify, track and apprehend suspects, but also has wider uses, e.g., in education and theatre.

It is instructive to consider, as the Court omitted to do, whether the claimed subject matter falls positively within the "process" category of § 101.

The representative claim concerns a method for creating a composite image although it expressly recites only a single facial feature.  It starts with the steps of displaying facial feature images on a first area of a first display, the facial feature images being associated with facial feature element codes.  A facial feature image from the first area of the first display is selected via a user interface of the first device.  These steps are illustrated by the accompanying image downloaded from the website.  The user is offered a range of relatively small hairline images each one of which has an associated element code.  A selected hairline image is transferred from the first (right hand) area of the display to the second (left hand) area where a larger image is formed, a multiplication step being performed to show the selected hairline on an enlarged scale and to derive a facial image code.

However, as explained, e.g., with reference to Fig. 4 of the patent, to complete the composite image further classes of feature image are added to provide a complete facial image code which is then displayed on a second display, these further steps being illustrated by the following further images from the website.

The criteria for eligibility under the "process" category of § 101 are set out in Gottschalk v. Benson (1972), quoting Justice Bradley in Cochrane v Deener (1876):

A process is a mode of treatment of certain materials to produce a given result.  It is an act, or a series of acts, performed upon the subject matter to be transformed and reduced to a different state or thing.  If new and useful, it is just as patentable as is a piece of machinery.  In the language of the patent law, it is an art.  The machinery pointed out as suitable to perform the process may or may not be new or patentable; whilst the process itself may be altogether new, and produce an entirely new result. The process requires that certain things should be done with certain substances, and in a certain order; but the tools to be used in doing this may be of secondary consequence.

Novelty was not in question and utility is supported by the fact that FACES has received a number of awards including the President's Award, Crime Stoppers International, 1999 as an outstanding contribution to law enforcement in the field of crime solving and is an official training tool for FBI agents.  The website includes quotes from serving police officers, of which the following is representative:

"I have been a law enforcement officer for over 30 years, having worked a variety of assignments which required composite drawings of suspects.  For those drawings I utilized a system utilizing individual foils to construct the desired image.  A complex, time-consuming process under the best conditions.  After becoming familiar with your system, I was so impressed that I showed several officers how to develop a composite drawing using your FACES software.  I know that this valuable gift will enhance our efforts in bringing to justice those who are intent on violating the law at the expense of society."  Byron C. Nelson, Chief of Police, Sequim Police Dept., Sequim, Washington.

The subject matter to be transformed is the various facial feature images, and their conversion to a composite image and reproduction on a second display are plainly transformational within the test set out in Gottsschalk.  The new result is set out in the '303 patent in problem/solution terms familiar to European practitioners and was acknowledged in the panel opinion.

Prior to the invention disclosed in the '303 patent, composite facial images typically were stored in file formats such as "bitmap," "gif," or "jpeg."  But these file formats required significant memory, and compressing the images often resulted in decreased image quality.  Digital transmission of these images could be difficult.  The '303 patent sought to solve this problem by encoding the image at one end through a variety of image classes that required less memory and bandwidth, and at the other end decoding the images.

The commercial success and police and other acclaim for FACES provides substantial evidence to support the inventive merit of the subject matter claimed in the '303 patent, and is strikingly reminiscent of the commercial success and industry praise in the en banc decision of the Federal Circuit in Apple Inc. v. Samsumg Electronics (October 2016), where a video showed Steve Jobs unveiling the slide to unlock feature of an iPhone at an Apple event and that when he swiped to unlock the phone, "the audience burst into cheers."  It was held that the jury had been presented with substantial evidence of praise in the industry that specifically related to features of the claimed invention, thereby linking that industry praise with the patented invention.

It therefore appears almost beyond argument that the claimed subject matter indeed falls positively within the process category, and that this should have been a conclusive holding avoiding the need for any consideration of the ineligible "abstract idea" category.  It is questionable whether that category can be invoked to "nibble away" at the full scope of the "process" category without raising issues of separation of powers.  Shortly before his appointment to the Supreme Court, Judge Neil Gorsuch (as he then was) explained in Gutierrez-Brizuela v. Lynch (2016):

Conversely, what would happen if politically unresponsive and life-tenured judges were permitted to decide policy questions for the future or try to execute those policies?  The very idea of self-government would soon be at risk of withering to the point of pointlessness.  It was to avoid dangers like these, dangers the founders had studied and seen realized in their own time, that they pursued the separation of powers.  A government of diffused powers, they knew, is a government less capable of invading the liberties of the people.  See  The Federalist No. 47 (James Madison) ("No political truth is . . . stamped with the authority of more enlightened patrons of liberty" than the separation of powers).

Scalia and Garner in their classic book Reading Law, The Interpretation of Legal Texts, Thomson/West, 2012 devote an entire chapter to the so-called Omitted-Case Canon, supporting the view that supplying omissions in statute law transcends the judicial function, and quoting William Blackstone in the introduction:  "[L]aw, without equity, though hard and disagreeable, is much more desirable for the public good than equity without law: which would make every judge a legislator, and introduce most infinite confusion."  In the context of eligibility under § 101 the classic statement in United States v. Dubilier Condenser Corp that the Court should not read into the patent laws limitations and conditions that the legislature has not expressed has been cited with approval in Diamond v Chakrabarty and Diamond v Diehr and has never subsequently been questioned.

It is strongly arguable that the Alice test for excluded subject-matter was employed in Regniocorp with disregard for the statute and producing a result contrary to what statute intended.  It is submitted that decisions such as Prometheus, Alice, and Myriad should be interpreted in accordance with standards rules of interpretation and should not be gold-plated, especially given the clearly expressed and almost universally ignored caution with which those decisions were expressed.  In Alice, Justice Thomas indeed said that:  "We must first determine whether the claims at issue are directed to a patent-ineligible concept."  However, that remark is most plausibly interpreted as a pragmatic response to the facts and arguments with which the Court was presented and not as a general instruction as to the approach to be followed in all future cases.  It should not be interpreted as a decision by a side wind that the judicial exclusions should enjoy priority over the categories of eligibility created by Congress in § 101.

In the panel's Alice step one analysis, it was held that the claim was directed to the abstract idea of encoding and decoding image data and reflected standard encoding and decoding.  No references were quoted in support of that proposition, and the opinion does not record any evidence about what was standard practice immediately prior to the invention in the technical field of encoding and decoding image data.  The examples given are in different fields and for different purposes:  Morse code for the transmission of alphanumeric characters and Paul Revere's signalling system for the transmission of pre-arranged instructions, neither of which have relevance to the transmission of image data.  The panel fastened onto a reference in the Background section of the patent to paper strips containing exemplary features, without acknowledging the inconvenient truth that they were in hard-copy format and that there was no evidence that the idea of encoding those strips had ever occurred to users.  While acknowledging the warning in Enfish that an improperly high level of abstraction should be avoided and recited features should not be ignored, the panel concluded that "[the] patent claims a method whereby a user starts with data . . . and ends with a new form of data."  As clearly recited in the claim the user starts with facial feature images on a display, selects particular features to form a composite image, and ends with a composite image reproduced on a second display.  The allegation that the user starts with and ends with data is therefore wrong not merely as a matter of opinion but as a matter of plain fact, and accordingly, the holding that the claim is directed to subject-matter that is no more than an abstract idea is untenable.

In Alice step two, the panel held that "[h]ere, RecogniCorp has not alleged a particularized application of encoding and decoding image data.  Indeed, claim 1 does not even require a computer; the invention can be practiced verbally or with a telephone.  J.A. 28 (col. 4 ll. 59–63); J.A. 32 (col. 11 ll. 53–59)."  Although the claimed method reduces facial features to simple codes and magnifications that can be transmitted verbally by telephone (although this seems not to be the normal use in practice) the panel has picked on the verbal transmission blindly and without reference to its overall significance and context, and has magnified that reference out of all proportion.  In reality, the patent deals with displays (which as a matter of common general knowledge and context form parts of computers) and with images on those displays which as is the common experience of all of us who send images as e-mail attachments involve tens or hundreds of kilobytes of data.  The suggestion that "the invention can be practiced verbally" properly belongs in the fictional Wonderland of Lewis Carrol's Alice rather than the real world in which we are all compelled to live.  Moreover, the significance of "multiplication" has been arbitrarily disregarded whereas it is apparent in context that it relates to necessary scaling and image production and is a technical feature of the process.

It is understood that an application for en banc reconsideration is likely to be filed.  Not only is it difficult to identify a more deserving case for such reconsideration, but as a matter of public interest the opportunity should be taken to re-focus on statutory compliance (which as relevant cast-law shows already contains non-trivial tests) and to relegate the judicial exceptions to the exceptional cross-checking role that they correctly deserve.

* Mr. Cole is a European Patent Attorney and Partner with Lucas & Co. in Warlingham, Surrey, UK and Visiting Professor at the Centre for Intellectual Property Policy & Management at Bournemouth University.

Comments

[Dan Feigelson]

Nice analysis, Paul. But after Sequenom, it's clear that most CAFC judges and those on the SCOTUS couldn't care less about whether or not someone actually invented something - they know patent-ineligible things when they see them. Sadly, this en banc petition will gain no traction.

[EG]

Hey Paul,

It's not just Reginocorp that's a miscarriage of justice-it's the whole nonsensical and broken Mayo/Alice framework that is. And as one of the amicus briefs in from Europe in support of the failed Sequenom cert petition to SCOTUS astutely suggested, the US is HUGE danger of being deemed in violation of TRIPS because of this nonsensical and broken Mayo/Alice framework for determining patent-eligibility.

[Paul Cole]

@ Dan Felignson

If claimed subject matter is truly eligible and inventive, it is incumbent on the attorney for the patentee to get the facts and the law right and to point them out to the judges in the panel (who may be inexperienced in patent law) in a way that they can understand and appreciate. We should not complain about poor outcomes if they are the result of poor advocacy. If the law is correctly pointed out to the panel and then disregarded, that is another matter.

@ EG

Both the CIPA and the EPI briefs (of which I was lead author) pointed out the problems of TRIPS compliance. It is my personal view that Sequenom complied neither with US law as explained here nor with TRIPS. Chapter and verse is to be found in the CIPA brief, though the separation of powers issue was made implicit rather than being positively spelled out (which we thought we did not actually need to do, because we expected that the clerks in the Supreme Court would be legally educated enough to get the point).

One of the major points in this posting is that the courts should look first at positive Section 101 compliance, in which case the three non-eligible categories are needed only as a cross-check. But beware: each of the four eligible categories has its own case-law baggage and is not a push-over.

[Skeptical]

Bravo Mr. Cole.

Recently you indicated that your opinion (and hence treatment) of the courts in regards to 101 was undergoing some self-inspection.

It appears that a new level of disdain is arising, and I for one would say: rightly so.

[GrzeszDeL]

This essay is a gem, just like everything else that Paul Cole writes. Thanks for publishing it.

[Stare Decisis]

Paul,

Are you sure that you want to describe a process and an abstract idea as distinct categories? I don't know how it works in Europe, but in the US there's much jurisprudence, not the least of which was Alice, that a claim can be both directed to a process AND directed to an abstract idea. That's fundamental to the US concept of a judicial exception, as spelled out in at least Alice.

So your paragraph pointing out the problem with not having categorized the Recognicorp claim as a process seems to draw an incorrect conclusion. In Recognicorp, there was never any issue that the claim was a process, so yes that was assumed and not discussed explicitly, and then the claim also was found to be directed to an abstract idea. This is not the problem with Recognicorp, at least not per current US case law.

Two other points:

It sounds like you and the other commenters here are attacking the very concept of US judicial exceptions, that they judicially steal from the authorized statutory categories, which of course is a bigger issue than is addressed in Recognicorp. An argument along those lines really needs to address the US "Supreme Court Trilogy" of Gottschalk, Flook and Diehr in their entirety. That's a much bigger lift.

Also, I agree that Recognicorp is flawed in that its holding ignores an improvement analogous to McRO which is clearly stated right in the background section of the opinion. That's a more fact-specific issue directly relevant to Recognicorp which it would have been helpful for the court to have addressed, consistent with Enfish, McRO, etc.

[Paul Cole]

@ Stare Decisis

In my view the eligible categories and the judicial exceptions at least in their original definitions are distinct non-intersecting sets if viewed on a Venn diagram.

The classic abstract idea in the present field is a mathematical algorithm or an improved method of calculation as in Parker v Flook. Arguably these are insufficiently transformative to count as processes, compare the classic EPO Appeal Board opinion in T 208/84 Computer-related invention/VICOM where the original claim to a method of filtering a two-dimensional data array was considered ineligible, but an amended claim to a method of digitally processing images in the form of a two-dimensional data array was allowed. The Board concluded that: "A basic difference between a mathematical method and a technical process can be seen, however, in the fact that a mathematical method or a mathematical algorithm is carried out on numbers (whatever these numbers may represent) and provides a result also in numerical form, the mathematical method or algorithm being only an abstract concept prescribing how to operate on the numbers. No direct technical result is produced by the method as such. In contrast thereto, if a mathematical method is used in a technical process, that process is carried out on a physical entity (which may be a material object but equally an image stored as an electric signal) by some technical means implementing the method and provides as its result a certain change in that entity. The technical means might include a computer comprising suitable hardware or an appropriately programmed general purpose computer."

It occurs to me that the present decision, having regard to established EPO policy since VICOM, additionally raises issues of TRIPS compliance.

There is a difference between a set of steps merely labelled as a process and a set of steps which in reality as a matter of substance falls into the eligible "process" category. It is like Abraham Lincoln's story about the legs on a dog - merely calling the tail a leg does not make it into a fifth leg for the dog. My submission is that if subject matter falls in reality into an eligible category, then the judicial exceptions cannot apply.

I am not attacking the concept of the judicial exceptions but only their over-expansive interpretation to cover subject-matter that congress intended should be allowable. This is, of course, a serious issue and the Sequenom case was described to me by a senior and knowledgeable individual when it was still before the Supreme Court as "dynamite".

[Stare Decisis]

Paul,

Thanks for the additional thoughts. I'd also like to offer some further thoughts, at least regarding prosecution in the US. Here is not so much a rebuttal as a different perspective / framing.

In some ways I agree that it seems a somewhat semantic point whether judicial exceptions (JEs) are internally or externally distinct from the 101 statutory categories. Either way, they subtract from the patent eligible claims, possibly too much so depending on JE case law, as evidenced by Recognicorp.

However, I do think that care must be taken in arguing that a claim is not directed to a JE because it is fits the definition of a 101 statutory category. This mixes the tests and relevant case law, since, at least in US practice, a claim can indeed fit a case-law-derived definition of a textual 101 statutory category and still be directed to a judicial exception.

Put another way, using the test for 101 statutory categories, there are many reasons why a claim may fail to fit into a textual category (some examples are listed in US MPEP 2106.I, e.g. transitory forms of signal transmission and mere arrangement of printed matter). However, that a claim passes the relevant tests and is properly directed to a 101 category says nothing about whether it passes the still evolving JE tests and also is directed to a JE. Again, different tests. All but one of the MPEP examples turn on simple physicality, but US 101 case law has evolved into a more complex, and currently poorly defined, set of tests. To make a US analogy, 101 JE analysis has almost evolved into a framework with the complexity of the dreaded Erie doctrine Choice of Law analysis familiar to some only from law school, complete with necessary flow charts. But, like it or not, that is the situation, and it seems, at least for prosecution purposes, communication with examiners is facilitated by accepting the current system, for now at least, and working to move quickly through the "flow chart."

After listing evidence of commercial success, your article stated "It therefore appears almost beyond argument that the claimed subject matter indeed falls positively within the process category...," which I think most everyone would agree to, according to the relevant tests, including the Recognicorp panel. But then your article states therefore "...this should have been a conclusive holding avoiding the need for any consideration of the ineligible 'abstract idea' category," which, at least in form, according to US case law, appears to mix 101 tests. I know that in practice, the tests of whether a claim fits a textual 101 category and even whether an alleged JE is present are both quickly over and done with, and the most productive arguments relate to either highlighting non-conventional additional elements or improvement, either of which draws on entirely different case law from those initial questions.

Re-reading your article, it seems to me that many of the points in the first half of the article go to the question of improvement as demonstrated by commercial success. On that topic, the courts have opened a door through 101 via Enfish, McRO, etc. (unfortunately a door which the Recognicorp panel chose not to explore).

In that sense, I couldn't agree more with your points. Not only did the panel ignore their own background statement that the invention resulted in at least some form of improvement, but they did not give any consideration to the evidence for improvement which you highlight, and this despite authoring judge Reyna having played a role in recent cases highlighting the value of considering improvement, e.g. McRO. This, however, may go to the benefit of getting improvement clearly and overwhelmingly on the record.

So, as a practical point regarding at least prosecution, since the USPTO must synthesize and work with US case law as it exists, for now at least, it seems more productive to stay within that framework, moving past identification of a 101 category and even past identification of at least an alleged JE, to focus as you have done on how a claim is applied to the real world and what benefits result, i.e. improvement as something significantly more. Then, if the combination of additional elements is merely conventional, and if the novel / "non-conventional" part of a claim lies in the JE, effort can be focused on making it clear on the record how the JE is applied to result in an improvement and how that demonstrates the necessary something significantly more to satisfy 101.

Thus, if as you suggest Recognicorp does stretch the boundaries of the abstract idea JE too far, that can be framed as having given too little weight to the evidence of improvement: in the record, even in the opinion, and in your article -- or to the record having been insufficiently clear regarding improvement.

While preemption supposedly lies at the heart of the JEs, in recent US practice it seems more that application is the key to avoiding preemption, and clear improvement is recently the best way to demonstrate application, with the understanding that what's relevant is improvement in application as opposed to an improved algorithm in the "abstract."

In this regard, in practice, better improvement arguments are more specific and possibly evidence-based, as examples: a sufficient simplification of and acceleration in the "complex, time-consuming process" of Recognicorp, or an improved true positive rate relating to some output of a claim, or a measurable reduction in resources required. Such specific explanations have the possibility of going further in making it clear that a claim does result in a "particularized application" ( to use the Recognicorp phrase), as demonstrated by sufficiently on the record improvement.

If the Recognicorp panel did consider the improvement on the record but found it lacking, it is unfortunate that this was not addressed in more detail, since the case could have yielded a useful standard (i.e. negative example) regarding what form and amount of improvement is needed, or maybe such analysis, like yours, would even have led the court to find the claim eligible.

Thanks very much for the article and discussion.

[Paul Cole]

May I just take this opportunity to thank the commentators, and especially EG and Stare Decisis for a most stimulating discussion. It should be noted that comments like those of Stare Decisis do not get written in 5 minutes, but are the result of much prior knowledge, effort and thought. So the thanks are truly well-deserved.

[Kevin E Noonan]

This thread illustrates one of the reasons we started this blog. Bravo and thanks.