[IPOL discuss] optic flow algorithms in ipol

[Jean-Michel Morel]

Dear Juan and all,

It's fun that just in the same day we have three different persons 
mentioning their intention to publish Horn & Schunck in IPOL (Enric, 
Daniel, and this your colleague). This is precisely why I have proposed 
to start a list of publication projects, to avoid duplication and foster 
collaboration.  Nevertheless nothing opposes two different publications 
of the same algorithm at IPOL, provided of course they are for some good 
reason substantially different. Otherwise, the best thing is to make 
joint papers.  Yes, definitely we should announce all (serious) projects 
to avoid overlap.

Thus in answer to your  questions:

  So I will recommend him to publish the second option if
nobody has objections.

----->>>> O.K.

In addition, it could be useful to compare different implementations
of the same algorithm, but I don't how can be organize that.
Should we make one publication for the Horn & Schunck algorithm with
many sub-parts devoted to each implementation and links to many demos?
Should we compare the implementations internally and pick one (or
two)?
What do you think?

------>>>> Unless somebody takes up the coordination it will be the 
first in first out rule: whoever submits first and is accepted publishes 
the algorithm. The second submission will have to demonstrate that there 
is something substantially new or different in the interpretation of the 
algorithm, and so on. Of course if the submissions are simultaneous as 
editors we will try to make authors interact.

In the near future we will be contributing with a few more algorithms,
should I announce them here to avoid overlap? Or we are going to
implement the idea of Prof. Morel of making a centralized list of
works in progress?

------->>>>  Yes, please, can someone implement this, which only means 
some open list in which everyone can place a parcimonious description of 
their current projects.

Jean-Michel






Juan Cardelino a écrit :
> Dear all,
>             We have an author from one of our "friend" labs here in
> latin america, which want to publish an optic flow algorithm in IPOL.
> He has implemented these two:
> 
> Optical flow, Horn & Schunck 1981
> http://www.sciencedirect.com/science/article/pii/0004370281900242
> 
> Combined Local-Global, Bruhn & cols. 2005
> http://www.springerlink.com/content/t32m75w616838671/
> 
>  and he is willing to contribute them to IPOL. I've seen that Enric
> has implemented a comparison of many implementations, including Horn
> and Schunck. So I will recommend him to publish the second option if
> nobody has objections.
> In addition, it could be useful to compare different implementations
> of the same algorithm, but I don't how can be organize that.
> Should we make one publication for the Horn & Schunck algorithm with
> many sub-parts devoted to each implementation and links to many demos?
> Should we compare the implementations internally and pick one (or
> two)?
> What do you think?
> 
> In the near future we will be contributing with a few more algorithms,
> should I announce them here to avoid overlap? Or we are going to
> implement the idea of Prof. Morel of making a centralized list of
> works in progress?
> 
> Best regards,
>                    Juan
> _______________________________________________
> discuss mailing list
> discuss at list.ipol.im
> http://tools.ipol.im/mailman/listinfo/discuss
>