[vtkusers] SIFT legal issues (PATENTS)

David Doria daviddoria+vtk at gmail.com
Wed Dec 9 11:54:32 EST 2009


On Tue, Dec 8, 2009 at 7:03 PM, Luis Ibanez <luis.ibanez at kitware.com> wrote:
> David,
>
>
> A U.S. Patent grants a 20-years long Monopoly
> on the following rights for an invention:
>
>      * Use
>      * Make
>      * Sell
>      * Offer for sale
>      * Import
>
>
> These rights are "Exclusive" rights, meaning that the
> patent holder (the "assignee") does NOT necesarily
> have the right to exercise any of these actions, but she
> has the right to *exclude* others from performing these
> actions.
>
>
> The reason why the assignee doesn't necessarily have
> the right to perform these actions is that they may require
> for her to acquire licenses from other patent holders on
> pre-requisites for this particular invention. (e.g. you could
> have the patent for the Light-Saber, but I may have the patent
> for "Methods and apparatus for Jedi Fighting", so you still
> need to get a license from me, before you start making,
> using, selling, offering for sale... your Light-Sabers...).
>
> You also have to get a license from the patent holder of
> "Futuristics weapons" and the patent holder of "Plasma-based
> weapons, and the patent holder of "Apparatus activated by
> The Force"... )
>
>
> This 20-year long Monopoly (that is a Patent) is granted
> *in exchange* for the full public disclosure of the invention.
> In principle the patent document is such public disclosure,
> and that document is NOT copyrighted.
>
>
> Talking, writing, commenting, criticizing, reviewing, and teaching
> about Patented inventions are all rights that everybody have.
>
>
> Allowing for such dissemination of information is actually the
> other half of the "Economic bargain" that the US Congress offers
> as an incentive for innovators when it gives them the Patent
> Monopoly that allows them to exclude all of the 300 million
> people in the US to use those methods for the next 20 years.
>
>
> That dissemination is one of the purposes of giving the patent
> holder that 20-year long Monopoly, despite the fact that is well
> known that Monopolies are destructive for the Market Economy
> (because they simply prevent competition).
>
>
> You certainly have the right to publish papers about Patented
> methods (whether they are your Patents, or someone else's
> patents).
>
>
> What is very dangerous and misleading is when authors publish
> papers WITHOUT disclosing the fact that the methods presented
> in a paper are the subject of a Patent.  In those cases, readers
> may make the assumption that the methods are in the Public
> Domain and they are available for anyone to use (make, sell,
> offer for sale or import...).  Readers then proceed to start using,
> such methods and to become dependent on them, only to
> discover years later that somebody had acquire a Monopoly
> on their use, and that they are requesting a Court to order you
> to stop using "their" method, and to pay them for the damages
> that you have inflicted to their economic market.
>
> This is usually called "Submarine Patents" and IMHO are a
> display of unethical behavior on the part of the authors of such
> papers.  A classical example is the LZW patents + papers and
> their implications for the use of GIF images on the Internet.
>
>
> Unfortunately, many technical Journals and Conferences do NOT
> have a policy of Patent disclosure. (Medical conferences, on the
> other hand, do this on a routine basis. e.g. if you go to RSNA to
> claim that Light-Sabers are the cure for liver tumors, they will ask
> you to disclose if you have financial interests on any company
> that produces Light-Sabers, because, that obviously will bias your
> opinion on the effectiveness of   "Jedi-based tumor treatment
> methods").
>
> See for example:
> http://www.rsna.org/publications/rad/PIA/policies/conflicts.html
>
>
> So,... you are very welcome to submit papers about Patented
> methods to the Insight Journal (whether you hold such Patents
> or somebody else does). However, I would strongly encourage
> you to *disclose* in the first sentence of the abstract and in the
> first sentence of the Introduction, that one (or several) aspects
> of the methods described in the paper are covered by US (or
> other countries) Patents.
>
>
> A typical disclosure could look like:
>
>    "Portions of the methods described in this paper are
>      covered by US. Patents 6,666,666 and 6,999,666."
>
>
> The Patent number is sufficient for anybody to find the text
> of the patent at the USPTO web site (http://www.uspto.gov/)
> or at Google Patents (http://www.google.com/patents).
>
>
> With such disclosure, readers can make an *informed*  decision
> on whether they want to continue reading the paper and maybe
> trying the said methods, and maybe get a license from the patent
> holder.
>
> It is also good to warn readers about the fact that when they
> build software that implements such patented methods they
> are "making" the invention and therefore they are infringing
> on the Patent.  When they "run" the software, they are "using"
> the invention, and again, they are infringing on the Patent.
>
>
> Use in "Research" is not really exempt from the restrictions
> of the Patent.
>
>
> There are only two types of "Research" activities for which use
> of Patented inventions without permission, have been accepted
> as allowed:
>
>
> 1) When the Patent pertains to a Pharmaceutical,
>    and the research in question relates to the process
>    of studying the compound for the purpose of pursuing
>    FDA approval.
>
>
> 2) When you are doing it:
>
>               "Solely for amusement,
>                 to satisfy idle curiosity,
>                 or for strictly philosophical inquiry,"
>
> http://cyber.law.harvard.edu/people/tfisher/2002Madeyedit.html
> http://www.the-scientist.com/article/display/13615/
>
>
> ---
>
> If you are a member of any technical society that organizes
> Journals and Conferences (e.g. IEEE, ACM, ...) you should
> promote the practice of disclosure of Patented methods in
> all publications as a requirement for accepting a paper.
>
> As a reader of their papers you have the right to know if
> they are being published with a marketing intent.
>
>
> Authors who do not disclose Patents are engaging in a de-
> ceptive behavior, and simply using Conferences and Journals
> as an inexpensive marketing platform for preparing the market
> of their prospective products.
>
>
> ----
>
> This is particularly damaging in our technical field,
> because there are today about 170,000 Software
> related patents in the USPTO.
>
> and some of them cover basic topics such as:
>
> 1)  pop-up windows in a browser
> 2)  Fourier Transforms
> 3)  Singular Value Decomposition
> 4)  Solving Linear systems
> ....
>
>
> Here is a reading short list:
>
> 5,835,392 Method for performing complex fast Fourier transforms
> 5,886,908 Method of efficient Gradient Computation
> 6,055,556 Apparatus and method for Matrix Multiplication
> 6,078,938 Method and system for Solving Linear Systems
> 6,356,926 Device and method for calculating FFT
> 6,434,582 Cosine algorithm for relatively small angles
> 6,640,237 Method and system for generating a Trigonometric Function
> 6,665,697 Fourier Analysis method and apparatus calculating
>                   the Fourier Factor Wn utilizing trigonometric relations
> 6,745,215 Computer apparatus, program and method for determining
>                   the equivalence of two algebraic functions
> 6,792,569 Root solver and associated method for solving finite field
>                   polynomial equations.
> 5,132,992 Audio and video transmission and receiving systems
>
> .....
>
> If you are not depressed yet,
>
> you can continue with the following patents
> on essential Web technologies:
>
>                   http://webshop.ffii.org/
>
>
> This is from the web site of the
>
>      "Foundation for a Free Information Infrastructure"
>
> one of the organizations that have opposed Software Patents,
> particularly in the European Union.
>
> --
>
>
> You may find interesting to read the following books:
>
>
> A)  "The Public Domain"
>      "Enclosing the Commons of the  Mind"
>        by James Boyle
>        Yale University Press
>       http://www.thepublicdomain.org/
>
>
> B)  "Patent Failure"
>      "How Judges, Bureaucrats,
>       and Lawyers Put Innovators at Risk"
>       by James Bessen and Michael J. Meurer
>       Princeton University Press
> http://www.researchoninnovation.org/dopatentswork/
>
>       Particularly this chapter on Software Patents:
> http://www.researchoninnovation.org/dopatentswork/dopat9.pdf
>
>
> C)   "Intellectual Property and Open  Source"
>       "A Practical Guide to Protecting Code"
>        by Van Lindberg
>        O'Reilly Press
> http://oreilly.com/catalog/9780596517960
>
>
> D)  "Math You Can't Use"
>      "Patents, Copyright and Software"
>       by Ben Klemens
>       Brookings Institution Press
> http://www.brookings.edu/press/Books/2005/mathyoucantuse.aspx
>
>
> E)  "Introduction to Economic Analysis."
>      McAfee, R. Preston, , and Tracy R. Lewis. 1969 .
>      Flat World Knowledge.
>      <http://www.flatworldknowledge.com/node/29467> .
>      (under Creative Commons License).
>
>
>
> You may also find interesting the material
> from our Lecture on "Software Patents":
>
> http://public.kitware.com/OpenSourceSoftwarePractice/images/3/33/OpenSourceSoftware-Patents-Fall2009.odp
>
> from the "Open Source Software Practices"
> course at RPI.
>
>
>
> The topics of Copyright and Patents should
> be mandatory material for anybody how is
> pursuing an Engineering degree.
> (or a Msc, or a Ph.D.)
>
>
>
> I would leave you with this quote:
>
> -------------------------------------------------------------
>
>   “Governor Thomas was so pleased with the
>    construction of this stove...
>    that he offered to give me a patent for the
>    sole vending of them for a term of years;
>    but I declined it from a principle which has
>    ever weighted with me on such occasions,
>
>    viz.:
>    That as we enjoy great advantages from the
>     inventions of others, we should be glad of
>     an opportunity to serve others by any
>     invention of ours; and this we should do
>     freely and generously.”
>
>                                  Benjamin Franklin
>
> ------------------------------------------------------------
>
>
>
>    Regards,
>
>
>          Luis
>
>
>
> ------------------------------------------------------------
> On Tue, Dec 8, 2009 at 4:16 PM, Bill Lorensen <bill.lorensen at gmail.com> wrote:
>> David,
>>
>> I think you can certainly publish. You have to be careful about how
>> you use it. Just because you are at a University does not give you
>> rights to use it as far as I know.
>>
>> Perhaps Luis will comment. Although he is not a lawyer, he knows much
>> more about this than I.
>>
>> Bill
>>
>> On Tue, Dec 8, 2009 at 3:59 PM, David Doria <daviddoria+vtk at gmail.com> wrote:
>>> On Tue, Dec 8, 2009 at 3:49 PM, Bill Lorensen <bill.lorensen at gmail.com> wrote:
>>>> David,
>>>>
>>>> According to wikipedia:http://en.wikipedia.org/wiki/Scale-invariant_feature_transform
>>>>
>>>>  "The algorithm is patented in the US; the owner is the University of
>>>> British Columbia.".
>>>>
>>>>  If this is true then we cannot include it in VTK or ITK.
>>>>
>>>> And you should be careful how you use it in your own work. You may be
>>>> violating the patent.
>>>>
>>>> Bill
>>>
>>> I thought that was the case about I/VTK.
>>>
>>> However, what about the VJ?
>>>
>>> I suppose I need a crash course in patent law, but am I really not
>>> allowed to even use it in my own work? Surely people have published
>>> papers saying "we then use SIFT to blah blah", right?
>>>
>>> Thanks,
>>>
>>> David

Luis,

As usual, an impressive, informative, and comprehensive response, thanks!

Am I correct in the summary that "Yes, you can submit code of patented
algorithms to the VTK Journal. It does not become illegal until the
code is compiled."?

So when people say "we used SIFT..." in a journal paper, they are
using the executable provided by Lowe. If they had compiled their own
implementation, they would be breaking the law. Correct?

Is it fair to say that these types of algorithms are still nice to
have in ITK/VTK format for those users who wish to use it for
"amusement only"?

Thanks,

David



More information about the vtkusers mailing list