I've just updated the site to run off Drupal 6 now. At first everything may seem a bit broken, so if you happen upon this site and it all looks broken, please come back in a few days time!

Joe

Using gluUnproject is perfectly adequate for telling where a mouse click went in a 3D world, but when you have a large number of depth points that you want to transform (like when you are using the depth map of a mesh to perform visibility calculations) then it is rather slow and unwieldy.

Wikipedia is a fantastic resource for maths and science. I've never yet got any bad information off the site in those fields (celebrity gossip however can be a bit more dubious). However I would never cite it, and sometimes I worry abbout the accuracy of some of the articles. And it seems I'm not the only one with these thoughts.

I am starting to have a look at SIFT for feature detection in images (well, what else are you going to use it for) and thought I would share with you my experiences of bringing the worlds of SIFT and Python together

Original Flavour SIFT

I'm shortly going to be restructuring this site, removing the CVSSP centric stuff (that didn't seem to work anyway) and opening things up a lot more. In the meantime I am searching around the net looking for other computer vision bloggers and trying to compile some sort of list of what is out there.

So first off is solem's vision blog someone else out there who is doing computer vision using numpy and python. A quick scan over his older posts has already sped up numpy on my PC by up to 4x so I'm already a fan!

I've just set up eyepy.org to redirect to here.

Eventually I hope to take some of the stuff I have been writing and put it all together as a python based computer vision package, possibly to go in as part of scipy, or possibly at a higher level bringing in openscenegraph compatability and some other bits and bobs.

Anyway, the first step (as ever) was to get a good name!

Joe

I was just reviewing my thesis draft and was thinking that my literature review on the topics of human pose and motion estimation were rather weak. Luckilly, someone over at IJCV seems to have heard me and they have put out a special issue on just that topic.

And what's more they have put together an editorial specifically calling for more research in some of the areas I've been looking at!

The Universe must love me.

Joe

I have just started reading the rather well put together blog over at Augmented Reality Times  where they have just posted a nice little roundup of some virtual makeup applications. It's interesting considering how face recognition really doesn't work in any general sense, how some specific applications are still being found that actually make useful products - like virtual makeup, and face recognition in modern cameras being used to guide auto-focus.

Well the site has been down for a while, but it's back again. There are some broken bits that will need fixing, but fingers crossed it shouldn't take too long to get everything working again...

Joe

Well, it's been a year since I purchased this domain, and I have a site and some users, but no one is really using this resource yet. Partly that's my own fault, I haven't had enough time to get the ball rolling properly, but partly I think it's because people want to use it for internal CVSSP stuff which means I have to lock it down and so it looks closed to the rest of the world.

So I now have a question I need to answer. Should I