[Robotics] GSoC 2013 mentor proposal (copy of improved application and additional info)

[Ricardo Fabbri]

Dear Colleagues from the Blender developer community,

I am posting this here since there is a large overlap between this
interest group and my interests with helping out with Blender
development.

I'd like to offer my voluntary service as a mentor during GSoC 2013. I
am a tenured professor at the State University of Rio de Janeiro, have
extensive experience in computer graphics in top players in the
industry and academia. My specialty is computer vision and image
processing, specially 3D reconstruction (structure from motion, match
moving, multiple view stereo, motion capture, etc). Prior to my
current permanent position as a professor I have interned and worked
full-time at Google in the US, and hold a PhD from Brown University.

I was an official mentor for GSoC 2012 for the Scilab org, having
helped 6 students from my university and hacker org to get accepted.
This year I am foreseeing many GSoC applications, trying to encourage
more students to apply than last year. In particular, I have
encouraged a very bright graduate student of my university to apply to
Blender, Marcos Couto (IRC handle ocf). Perhaps I could help Blender
either through direct mentoring of Marcos, or by mentoring someone
else but also helping out with Marcos, as I have direct contact with
him daily. My personal interests with this mentoring would be to use
the oportunity as a stimulus to develop match moving applications in
Blender, as well as video editing. I am directly interested in the
general software engineering that go into this scale of development of
this central computer graphics application. I have learned many things
at this scale with VTK, VXL (vxl.sf.net), Scilab, and working at
Google, but Blender surely complements that in its particularities and
in my desire to develop it for my research. I hope that as an
experienced programmer and computer graphics / computer vision
researcher I can somehow be useful to the Blender community in GSoC
2013. If so, I can easily and promptly get started with the
technicalities.

Regardless of GSoC, I personally plan to develop general inverse
rendering functionality in blender, match moving applications (3D
reconstruction of scene and object geometry from video) and the video
editor. As far as robotics go, I am interested in working with
flexible matchmoving systems that do not require precise prior
knowledge of camera parameters and that can work in general
uncontrolled environments.

If you have any doubts of my competence for this role you can use the
next few days to interact with me before accepting to connect here.
Best regards, Ricardo Fabbri.

PS: I apologize but in melange there is a wrong mention to LibreOffice
in my melange application, I meant Blender throughout. I typed
'Libreoffice' by complete accident as I was helping out another
student with his application there. Blender is clearly within my
expertise field, so please consider this application carefully as it
can lead to fruitful activity. I have funding incentives to develop
advanced functionality for video-related software outside of GSoC, and
have also worked with Kdenlive/MLT and with projects for writing a
Rivendell video application (which would be a TV/video broadcasting
application with added computer vision features to index video).
Moreover, I also authored of the Scilab image processing toolbox. I
hope this makes it clear that this is not a generic application, but
that I have solid interests in Blender.

Let me know if I can be of any specific help.
Ricardo Fabbri.

--
Dr Ricardo Fabbri
Professor of Computer Engineering
GNU/Linux registered user #175401
www.lems.brown.edu/~rfabbri
pt.wikipedia.org/wiki/IPRJ
labmacambira.sf.net