Hi everyone,

I have a great new patentable idea for image recognition. The idea is, given a representative selection of photographs of an entity that allow it to be essentially "envisioned" in 3 dimensions, we can decide if photographs might be containing an entity or type of entity based on certain math advancement ideas I've thought of. The tech would be particularly useful for face recognition.

-Philip White (philipjwhite@yahoo.com)

On 7/1/2022 2:06 PM, B.H. wrote:
>
> Hi everyone,
>
> I have a great new patentable idea for image recognition. The idea is, given a representative selection of photographs of an entity that allow it to be essentially "envisioned" in 3 dimensions, we can decide if photographs might be containing an entity or type of entity based on certain math advancement ideas I've thought of. The tech would be particularly useful for face recognition.

So, get a patent. It's really easy: the patent office barley checks
technical contents. And when I say easy one must remember that even PO
(yes, the fool who posts the most trash in this form) has bragged he got
a software patent. His only mistake was posting the application online
so we could all read it. The ideas where well and better thought out in
the 1960s!!! And most of those ideas became trash quickly. And if you
can't do better than PO, I'll lose all respect for you.

Your program: Write patent application; receive feedback; edit and
resubmit; have nice glass of ice tea and announce victory when your
patent is granted; find buyer; halt.
--
Jeff Barnett

On 7/1/2022 3:06 PM, B.H. wrote:
>
> Hi everyone,
>
> I have a great new patentable idea for image recognition. The idea is, given a representative selection of photographs of an entity that allow it to be essentially "envisioned" in 3 dimensions, we can decide if photographs might be containing an entity or type of entity based on certain math advancement ideas I've thought of. The tech would be particularly useful for face recognition.
>
> -Philip White (philipjwhite@yahoo.com)

Patent It Yourself: Your Step-by-Step Guide to Filing at the U.S. Patent
Office https://www.amazon.com/s?k=patent-it-yourself

It a very useful book and the MPEP manual for patent examining procedure
is online https://www.uspto.gov/web/offices/pac/mpep/index.html

The is the rule book that all the patent examiners use it is written in
ordinary English with no legalese because patent examiners are not lawyers.

The most important part of the patent is the claims. Unless you have a
brand new and useful combination of ideas the patent is supposed to be
refused.

--
Copyright 2022 Pete Olcott "Talent hits a target no one else can hit;
Genius hits a target no one else can see." Arthur Schopenhauer

On 7/1/2022 7:24 PM, olcott wrote:
> On 7/1/2022 3:06 PM, B.H. wrote:
>>
>> Hi everyone,
>>
>> I have a great new patentable idea for image recognition.  The idea
>> is, given a representative selection of photographs of an entity that
>> allow it to be essentially "envisioned" in 3 dimensions, we can decide
>> if photographs might be containing an entity or type of entity based
>> on certain math advancement ideas I've thought of.  The tech would be
>> particularly useful for face recognition.
>>
>> -Philip White (philipjwhite@yahoo.com)
>
> Patent It Yourself: Your Step-by-Step Guide to Filing at the U.S. Patent
> Office https://www.amazon.com/s?k=patent-it-yourself
>
> It a very useful book and the MPEP manual for patent examining procedure
> is online https://www.uspto.gov/web/offices/pac/mpep/index.html
>
> The is the rule book that all the patent examiners use it is written in
> ordinary English with no legalese because patent examiners are not lawyers.
>
> The most important part of the patent is the claims. Unless you have a
> brand new and useful combination of ideas the patent is supposed to be
> refused.
>

My patent recognizes machine generated character glyphs with 100%
accuracy using a DFA. In is meant to provide an automated interface
between applications that have no alternative interface.

--
Copyright 2022 Pete Olcott

"Talent hits a target no one else can hit;
Genius hits a target no one else can see."
Arthur Schopenhauer

On Fri, 1 Jul 2022 19:27:52 -0500
olcott <NoOne@NoWhere.com> wrote:

> On 7/1/2022 7:24 PM, olcott wrote:
> > On 7/1/2022 3:06 PM, B.H. wrote:
> >>
> >> Hi everyone,
> >>
> >> I have a great new patentable idea for image recognition.  The
> >> idea is, given a representative selection of photographs of an
> >> entity that allow it to be essentially "envisioned" in 3
> >> dimensions, we can decide if photographs might be containing an
> >> entity or type of entity based on certain math advancement ideas
> >> I've thought of.  The tech would be particularly useful for face
> >> recognition.
> >>
> >> -Philip White (philipjwhite@yahoo.com)
> >
> > Patent It Yourself: Your Step-by-Step Guide to Filing at the U.S.
> > Patent Office https://www.amazon.com/s?k=patent-it-yourself
> >
> > It a very useful book and the MPEP manual for patent examining
> > procedure is online
> > https://www.uspto.gov/web/offices/pac/mpep/index.html
> >
> > The is the rule book that all the patent examiners use it is
> > written in ordinary English with no legalese because patent
> > examiners are not lawyers.
> >
> > The most important part of the patent is the claims. Unless you
> > have a brand new and useful combination of ideas the patent is
> > supposed to be refused.
> >
>
> My patent recognizes machine generated character glyphs with 100%
> accuracy using a DFA. In is meant to provide an automated interface
> between applications that have no alternative interface.
100% accuracy?

What about a font where `l` and `1` lack serifs so look almost
identical?

/Flibble

On 7/2/2022 4:12 AM, Mr Flibble wrote:
> On Fri, 1 Jul 2022 19:27:52 -0500
> olcott <NoOne@NoWhere.com> wrote:
>
>> On 7/1/2022 7:24 PM, olcott wrote:
>>> On 7/1/2022 3:06 PM, B.H. wrote:
>>>>
>>>> Hi everyone,
>>>>
>>>> I have a great new patentable idea for image recognition.  The
>>>> idea is, given a representative selection of photographs of an
>>>> entity that allow it to be essentially "envisioned" in 3
>>>> dimensions, we can decide if photographs might be containing an
>>>> entity or type of entity based on certain math advancement ideas
>>>> I've thought of.  The tech would be particularly useful for face
>>>> recognition.
>>>>
>>>> -Philip White (philipjwhite@yahoo.com)
>>>
>>> Patent It Yourself: Your Step-by-Step Guide to Filing at the U.S.
>>> Patent Office https://www.amazon.com/s?k=patent-it-yourself
>>>
>>> It a very useful book and the MPEP manual for patent examining
>>> procedure is online
>>> https://www.uspto.gov/web/offices/pac/mpep/index.html
>>>
>>> The is the rule book that all the patent examiners use it is
>>> written in ordinary English with no legalese because patent
>>> examiners are not lawyers.
>>>
>>> The most important part of the patent is the claims. Unless you
>>> have a brand new and useful combination of ideas the patent is
>>> supposed to be refused.
>>>
>>
>> My patent recognizes machine generated character glyphs with 100%
>> accuracy using a DFA. In is meant to provide an automated interface
>> between applications that have no alternative interface.
>
> 100% accuracy?
>
> What about a font where `l` and `1` lack serifs so look almost
> identical?
>
> /Flibble
>

It exactly matches the pixels so only when they are 100% perfectly
identical does it have ambiguity. Also it uses a DFA so it processes a
whole screen in 1/10 of a second.

--
Copyright 2022 Pete Olcott

"Talent hits a target no one else can hit;
Genius hits a target no one else can see."
Arthur Schopenhauer

On 7/2/2022 3:12 AM, Mr Flibble wrote:
> On Fri, 1 Jul 2022 19:27:52 -0500
> olcott <NoOne@NoWhere.com> wrote:
>
>> On 7/1/2022 7:24 PM, olcott wrote:
>>> On 7/1/2022 3:06 PM, B.H. wrote:
>>>>
>>>> Hi everyone,
>>>>
>>>> I have a great new patentable idea for image recognition.  The
>>>> idea is, given a representative selection of photographs of an
>>>> entity that allow it to be essentially "envisioned" in 3
>>>> dimensions, we can decide if photographs might be containing an
>>>> entity or type of entity based on certain math advancement ideas
>>>> I've thought of.  The tech would be particularly useful for face
>>>> recognition.
>>>>
>>>> -Philip White (philipjwhite@yahoo.com)
>>>
>>> Patent It Yourself: Your Step-by-Step Guide to Filing at the U.S.
>>> Patent Office https://www.amazon.com/s?k=patent-it-yourself
>>>
>>> It a very useful book and the MPEP manual for patent examining
>>> procedure is online
>>> https://www.uspto.gov/web/offices/pac/mpep/index.html
>>>
>>> The is the rule book that all the patent examiners use it is
>>> written in ordinary English with no legalese because patent
>>> examiners are not lawyers.
>>>
>>> The most important part of the patent is the claims. Unless you
>>> have a brand new and useful combination of ideas the patent is
>>> supposed to be refused.
>>>
>>
>> My patent recognizes machine generated character glyphs with 100%
>> accuracy using a DFA. In is meant to provide an automated interface
>> between applications that have no alternative interface.
>
> 100% accuracy?
>
> What about a font where `l` and `1` lack serifs so look almost
> identical?
I've criticized PO many times about this dumb-shit patent. There are
several problems: 1) you must specify the font - this means that it
needs a pixel template for each character and size, 2) you must exactly
register the specimen document horizontally, vertically, by angle, etc.
so that there are zero match errors, 3) the whole thing was designed and
engineered by someone who can't program and thinks correct software
engineering decisions are made by being steal against comments on mistakes.
It's been a while since I actually read it. I do know that the whole
thing reads like a comedian's gag sheet. It was the final thing that
convinced me that the US Patent Office is completely off the rails with
software patents.
I remember the following in the mid 1960s: A RAND tablet that could read
the x/y coordinates of a stylus and project pixel images on its surface.
Software using that tablet where the user a user could write
mathematical equations doing such things as writing summations using a
large sigma, do subscripts and superscripts, and the rest of the usual
trimmings. The software would take the hand written input and in real
time project a neatened and usually accurate image of the intended
symbols. It actually produced (inside) a text string telling what it
thought it read. Think of a miniature piece of a TeX like language.
Hardware was a 48-bit 48K word memory machine running time sharing with
a typical load of 15 to 20 interactive users. It's memory cycle time was
in the order of a microsecond. The rest of the hardware was a PDP 1 used
as a front end and buffer between the RAND tablet and the time sharing
computer.
This combination worked well for many users who wrote characters some
what differently, at different angles and sizes, etc.
--
Jeff Barnett

On 7/2/2022 1:26 PM, Jeff Barnett wrote:
> On 7/2/2022 3:12 AM, Mr Flibble wrote:
>> On Fri, 1 Jul 2022 19:27:52 -0500
>> olcott <NoOne@NoWhere.com> wrote:
>>
>>> On 7/1/2022 7:24 PM, olcott wrote:
>>>> On 7/1/2022 3:06 PM, B.H. wrote:
>>>>>
>>>>> Hi everyone,
>>>>>
>>>>> I have a great new patentable idea for image recognition.  The
>>>>> idea is, given a representative selection of photographs of an
>>>>> entity that allow it to be essentially "envisioned" in 3
>>>>> dimensions, we can decide if photographs might be containing an
>>>>> entity or type of entity based on certain math advancement ideas
>>>>> I've thought of.  The tech would be particularly useful for face
>>>>> recognition.
>>>>>
>>>>> -Philip White (philipjwhite@yahoo.com)
>>>>
>>>> Patent It Yourself: Your Step-by-Step Guide to Filing at the U.S.
>>>> Patent Office https://www.amazon.com/s?k=patent-it-yourself
>>>>
>>>> It a very useful book and the MPEP manual for patent examining
>>>> procedure is online
>>>> https://www.uspto.gov/web/offices/pac/mpep/index.html
>>>>
>>>> The is the rule book that all the patent examiners use it is
>>>> written in ordinary English with no legalese because patent
>>>> examiners are not lawyers.
>>>>
>>>> The most important part of the patent is the claims. Unless you
>>>> have a brand new and useful combination of ideas the patent is
>>>> supposed to be refused.
>>>
>>> My patent recognizes machine generated character glyphs with 100%
>>> accuracy using a DFA. In is meant to provide an automated interface
>>> between applications that have no alternative interface.
>> 100% accuracy?
>>
>> What about a font where `l` and `1` lack serifs so look almost
>> identical?
> I've criticized PO many times about this dumb-shit patent. There are
> several problems: 1) you must specify the font - this means that it
> needs a pixel template for each character and size, 2) you must exactly
> register the specimen document horizontally, vertically, by angle, etc.
> so that there are zero match errors, 3) the whole thing was designed and
> engineered by someone who can't program and thinks correct software
> engineering decisions are made by being steal against comments on mistakes.
>
> It's been a while since I actually read it. I do know that the whole
> thing reads like a comedian's gag sheet. It was the final thing that
> convinced me that the US Patent Office is completely off the rails with
> software patents.
>
> I remember the following in the mid 1960s: A RAND tablet that could read
> the x/y coordinates of a stylus and project pixel images on its surface.
> Software using that tablet where the user a user could write
> mathematical equations doing such things as writing summations using a
> large sigma, do subscripts and superscripts, and the rest of the usual
> trimmings. The software would take the hand written input and in real
> time project a neatened and usually accurate image of the intended
> symbols. It actually produced (inside) a text string telling what it
> thought it read. Think of a miniature piece of a TeX like language.
>
> Hardware was a 48-bit 48K word memory machine running time sharing with
> a typical load of 15 to 20 interactive users. It's memory cycle time was
> in the order of a microsecond. The rest of the hardware was a PDP 1 used
> as a front end and buffer between the RAND tablet and the time sharing
> computer.
>
> This combination worked well for many users who wrote characters some
> what differently, at different angles and sizes, etc.

It makes perfect sense for its intended purpose.
Providing the means for an automated interface between user interfaces
so that one application can talk to another application without
requiring a human user to manually enter the data.

Conventional OCR takes 30 seconds to do what my system does in 1/10 of a
second and conventional OCR is stochastic rather than deterministic thus
inherently has less accuracy.

--
Copyright 2022 Pete Olcott

"Talent hits a target no one else can hit;
Genius hits a target no one else can see."
Arthur Schopenhauer

On 7/2/2022 12:39 PM, olcott wrote:
> On 7/2/2022 1:26 PM, Jeff Barnett wrote:
>> On 7/2/2022 3:12 AM, Mr Flibble wrote:
>>> On Fri, 1 Jul 2022 19:27:52 -0500
>>> olcott <NoOne@NoWhere.com> wrote:
>>>
>>>> On 7/1/2022 7:24 PM, olcott wrote:
>>>>> On 7/1/2022 3:06 PM, B.H. wrote:
>>>>>>
>>>>>> Hi everyone,
>>>>>>
>>>>>> I have a great new patentable idea for image recognition.  The
>>>>>> idea is, given a representative selection of photographs of an
>>>>>> entity that allow it to be essentially "envisioned" in 3
>>>>>> dimensions, we can decide if photographs might be containing an
>>>>>> entity or type of entity based on certain math advancement ideas
>>>>>> I've thought of.  The tech would be particularly useful for face
>>>>>> recognition.
>>>>>>
>>>>>> -Philip White (philipjwhite@yahoo.com)
>>>>>
>>>>> Patent It Yourself: Your Step-by-Step Guide to Filing at the U.S.
>>>>> Patent Office https://www.amazon.com/s?k=patent-it-yourself
>>>>>
>>>>> It a very useful book and the MPEP manual for patent examining
>>>>> procedure is online
>>>>> https://www.uspto.gov/web/offices/pac/mpep/index.html
>>>>>
>>>>> The is the rule book that all the patent examiners use it is
>>>>> written in ordinary English with no legalese because patent
>>>>> examiners are not lawyers.
>>>>>
>>>>> The most important part of the patent is the claims. Unless you
>>>>> have a brand new and useful combination of ideas the patent is
>>>>> supposed to be refused.
>>>>
>>>> My patent recognizes machine generated character glyphs with 100%
>>>> accuracy using a DFA. In is meant to provide an automated interface
>>>> between applications that have no alternative interface.
>>> 100% accuracy?
>>>
>>> What about a font where `l` and `1` lack serifs so look almost
>>> identical?
>> I've criticized PO many times about this dumb-shit patent. There are
>> several problems: 1) you must specify the font - this means that it
>> needs a pixel template for each character and size, 2) you must
>> exactly register the specimen document horizontally, vertically, by
>> angle, etc. so that there are zero match errors, 3) the whole thing
>> was designed and engineered by someone who can't program and thinks
>> correct software engineering decisions are made by being steal against
>> comments on mistakes.
>>
>> It's been a while since I actually read it. I do know that the whole
>> thing reads like a comedian's gag sheet. It was the final thing that
>> convinced me that the US Patent Office is completely off the rails
>> with software patents.
>>
>> I remember the following in the mid 1960s: A RAND tablet that could
>> read the x/y coordinates of a stylus and project pixel images on its
>> surface. Software using that tablet where the user a user could write
>> mathematical equations doing such things as writing summations using a
>> large sigma, do subscripts and superscripts, and the rest of the usual
>> trimmings. The software would take the hand written input and in real
>> time project a neatened and usually accurate image of the intended
>> symbols. It actually produced (inside) a text string telling what it
>> thought it read. Think of a miniature piece of a TeX like language.
>>
>> Hardware was a 48-bit 48K word memory machine running time sharing
>> with a typical load of 15 to 20 interactive users. It's memory cycle
>> time was in the order of a microsecond. The rest of the hardware was a
>> PDP 1 used as a front end and buffer between the RAND tablet and the
>> time sharing computer.
>>
>> This combination worked well for many users who wrote characters some
>> what differently, at different angles and sizes, etc.
>
> It makes perfect sense for its intended purpose.
> Providing the means for an automated interface between user interfaces
> so that one application can talk to another application without
> requiring a human user to manually enter the data.
It did work. A user could employ this technology to input parts of
programs. I believe those parts could be interpreted/compiled and handed
input or, if just an expression, evaluated. But since you seem
interested in speed: The technology I just described has less computer
power (a PDP 1 and ~5% of a "big" computer) than you'd expect to find in
your tooth brush. And like your tooth brush, it ran in real time.
> Conventional OCR takes 30 seconds to do what my system does in 1/10 of a
> second and conventional OCR is stochastic rather than deterministic thus
> inherently has less accuracy.
But you solved the wrong problem. According to you, you never sold a
nickel's worth of conversion services or the software. At a minimum,
that indicates the software engineering - the requirements and problem
definition phases - were off the rails.
And there is no reason what so ever to believe that 1) being
deterministic is more accurate than stochastic and 2) you know enough
mathematics or could learn it to try to justify that statement. In the
case of your application, things such as slant, smudges, creases, page
bending while scanning, and so on will need stochastic methods to make
intelligent guesses. In fact it would help to have a dictionary
(lexicon) and some grammar rules to make up for fouled input. I know of
zero applications that match YOUR need for perfect input in order to work.
--
Jeff Barnett

On 7/2/2022 3:04 PM, Jeff Barnett wrote:
> On 7/2/2022 12:39 PM, olcott wrote:
>> On 7/2/2022 1:26 PM, Jeff Barnett wrote:
>>> On 7/2/2022 3:12 AM, Mr Flibble wrote:
>>>> On Fri, 1 Jul 2022 19:27:52 -0500
>>>> olcott <NoOne@NoWhere.com> wrote:
>>>>
>>>>> On 7/1/2022 7:24 PM, olcott wrote:
>>>>>> On 7/1/2022 3:06 PM, B.H. wrote:
>>>>>>>
>>>>>>> Hi everyone,
>>>>>>>
>>>>>>> I have a great new patentable idea for image recognition.  The
>>>>>>> idea is, given a representative selection of photographs of an
>>>>>>> entity that allow it to be essentially "envisioned" in 3
>>>>>>> dimensions, we can decide if photographs might be containing an
>>>>>>> entity or type of entity based on certain math advancement ideas
>>>>>>> I've thought of.  The tech would be particularly useful for face
>>>>>>> recognition.
>>>>>>>
>>>>>>> -Philip White (philipjwhite@yahoo.com)
>>>>>>
>>>>>> Patent It Yourself: Your Step-by-Step Guide to Filing at the U.S.
>>>>>> Patent Office https://www.amazon.com/s?k=patent-it-yourself
>>>>>>
>>>>>> It a very useful book and the MPEP manual for patent examining
>>>>>> procedure is online
>>>>>> https://www.uspto.gov/web/offices/pac/mpep/index.html
>>>>>>
>>>>>> The is the rule book that all the patent examiners use it is
>>>>>> written in ordinary English with no legalese because patent
>>>>>> examiners are not lawyers.
>>>>>>
>>>>>> The most important part of the patent is the claims. Unless you
>>>>>> have a brand new and useful combination of ideas the patent is
>>>>>> supposed to be refused.
>>>>>
>>>>> My patent recognizes machine generated character glyphs with 100%
>>>>> accuracy using a DFA. In is meant to provide an automated interface
>>>>> between applications that have no alternative interface.
>>>> 100% accuracy?
>>>>
>>>> What about a font where `l` and `1` lack serifs so look almost
>>>> identical?
>>> I've criticized PO many times about this dumb-shit patent. There are
>>> several problems: 1) you must specify the font - this means that it
>>> needs a pixel template for each character and size, 2) you must
>>> exactly register the specimen document horizontally, vertically, by
>>> angle, etc. so that there are zero match errors, 3) the whole thing
>>> was designed and engineered by someone who can't program and thinks
>>> correct software engineering decisions are made by being steal
>>> against comments on mistakes.
>>>
>>> It's been a while since I actually read it. I do know that the whole
>>> thing reads like a comedian's gag sheet. It was the final thing that
>>> convinced me that the US Patent Office is completely off the rails
>>> with software patents.
>>>
>>> I remember the following in the mid 1960s: A RAND tablet that could
>>> read the x/y coordinates of a stylus and project pixel images on its
>>> surface. Software using that tablet where the user a user could write
>>> mathematical equations doing such things as writing summations using
>>> a large sigma, do subscripts and superscripts, and the rest of the
>>> usual trimmings. The software would take the hand written input and
>>> in real time project a neatened and usually accurate image of the
>>> intended symbols. It actually produced (inside) a text string telling
>>> what it thought it read. Think of a miniature piece of a TeX like
>>> language.
>>>
>>> Hardware was a 48-bit 48K word memory machine running time sharing
>>> with a typical load of 15 to 20 interactive users. It's memory cycle
>>> time was in the order of a microsecond. The rest of the hardware was
>>> a PDP 1 used as a front end and buffer between the RAND tablet and
>>> the time sharing computer.
>>>
>>> This combination worked well for many users who wrote characters some
>>> what differently, at different angles and sizes, etc.
>>
>> It makes perfect sense for its intended purpose.
>> Providing the means for an automated interface between user interfaces
>> so that one application can talk to another application without
>> requiring a human user to manually enter the data.
>
> It did work. A user could employ this technology to input parts of
> programs. I believe those parts could be interpreted/compiled and handed
> input or, if just an expression, evaluated. But since you seem
> interested in speed: The technology I just described has less computer
> power (a PDP 1 and ~5% of a "big" computer) than you'd expect to find in
> your tooth brush. And like your tooth brush, it ran in real time.
>
>> Conventional OCR takes 30 seconds to do what my system does in 1/10 of
>> a second and conventional OCR is stochastic rather than deterministic
>> thus inherently has less accuracy.
> But you solved the wrong problem. According to you, you never sold a
> nickel's worth of conversion services or the software. At a minimum,
> that indicates the software engineering - the requirements and problem
> definition phases - were off the rails.
>

I have a customer that I regularly met with that had his own technology
that would have greatly benefited from my technology.

Every time I spend years creating a patent as soon as I was done then
everything changed and obtaining the required samples became many-fold
more difficult.

> And there is no reason what so ever to believe that 1) being
> deterministic is more accurate than stochastic

It is a matter of estimating versus determining.

> and 2) you know enough
> mathematics or could learn it to try to justify that statement. In the
> case of your application, things such as slant, smudges, creases, page
> bending while scanning, and so on will need stochastic methods to make
> intelligent guesses.

Machine generated character glyphs that are directly generated on the
display screen are all exactly the same or have a finite set that are
all exactly the same. PDFs had three glyphs per character.

> In fact it would help to have a dictionary
> (lexicon) and some grammar rules to make up for fouled input. I know of
> zero applications that match YOUR need for perfect input in order to work.

--
Copyright 2022 Pete Olcott

"Talent hits a target no one else can hit;
Genius hits a target no one else can see."
Arthur Schopenhauer

On Friday, July 1, 2022 at 8:24:08 PM UTC-4, olcott wrote:
> On 7/1/2022 3:06 PM, B.H. wrote:
> >
> > Hi everyone,
> >
> > I have a great new patentable idea for image recognition. The idea is, given a representative selection of photographs of an entity that allow it to be essentially "envisioned" in 3 dimensions, we can decide if photographs might be containing an entity or type of entity based on certain math advancement ideas I've thought of. The tech would be particularly useful for face recognition.
> >
> > -Philip White (philip...@yahoo.com)
> Patent It Yourself: Your Step-by-Step Guide to Filing at the U.S. Patent
> Office https://www.amazon.com/s?k=patent-it-yourself
>

I have that one, it's really good.

> It a very useful book and the MPEP manual for patent examining procedure
> is online https://www.uspto.gov/web/offices/pac/mpep/index.html
>
> The is the rule book that all the patent examiners use it is written in
> ordinary English with no legalese because patent examiners are not lawyers.
>

I probably will get patents after I'm out of captivity; I don't want to risk weird legal issues while I'm still being stalked.

I realized that anything that could be patented as an electronic circuit could be patented as software, so Alice doesn't apply to disqualify many of my patents in the way I had initially thought. Probably, tons of my ideas are patentable and I'll get patents later and try to sell them.

> The most important part of the patent is the claims. Unless you have a
> brand new and useful combination of ideas the patent is supposed to be
> refused.
>

Thanks, that sounds like an important and helpful tip.

-Philip

> --
> Copyright 2022 Pete Olcott "Talent hits a target no one else can hit;
> Genius hits a target no one else can see." Arthur Schopenhauer