Right now the abstraction paper is working now after contacting with the author and doing a lot modifications.
The Stroke-based codes are still not working. I am thinking if it is okey to use their stroke models and use our placement to come up with the results. Otherwise, it is not working.
The codes for the paper 'image analogies' complied successfully, but cannot run due to unexpected error. If this one works, I guess it might replace the stroke-based paper. Now I am working on this too. But they said they won't support these codes anymore.
If needed, we can implement a new stroke placement algorithm. Or, we can use a different painterly image filter, if we can find one for which an implementation exists. Quite a nice algorithm was in CGI last year, maybe those guys have made their code public.
ReplyDeleteOh, and never mind image analogies. I am not surprised that it is unsupported.
ReplyDelete