Wikipedia:Reference desk/Archives/Computing/2011 August 26

= August 26 =

A little help please with new windows install/getting my hardware to work?
I have a Gateway LX6810-01 http://support.gateway.com/s/PC/LX/LX6810/LX6810sp3.shtml

i just installed windows 7 and started installing drivers from here: http://support.gateway.com/us/en/product/default.aspx?tab=1

...but its so confusing.... which ones do i get? So far ive installed....

The top (NVIDIA) audio driver... The bottom chipset driver... and i think the middle VGA driver

on that page.... Do i need any of the others, like the chipset driver that says its for "MCP7A" ?

Also, i have a 15 in one card reader that used to show up in "My computer" as 4 or 5 different devices, how do i get those to show again? Additionally i have a tv tuner and an infrared remote im wondering about as well.

I know its alot of items, but if anyone has knowledge on any of these hardware devices id very much appreciate it. Thanks!

216.173.144.164 (talk) 01:27, 26 August 2011 (UTC)


 * Did you try installing the detection utility from support page? Kushal (talk) 19:30, 31 August 2011 (UTC)

Missing audio channel?
Hello. I am listening to Pandora Radio and it seems that I am getting only one channel of stereo audio from the site. Sound is coming from both of my speakers, and my audio control is set to the middle, yet I only hear one channel of the music. There is also a bit of "echo" coming from Pandora (but not from any other site). Is there a setting somewhere that I've missed? — Michael J 04:49, 26 August 2011 (UTC)

extracting 3d coordiantes from 2d images of face
Is there a way that one can merge two or three 2d images of the same object (from different angles) to create its 3d model? Suppose i have a few samples of a face from a face database then i want to make its 3d model, the objective is to get the 3d coordinates of different points on the face. Can this be done or is there any alternative way to get the 3d coordinates?--tathagata 17:26, 26 August 2011 (UTC) — Preceding unsigned comment added by Nonstop funstop (talk • contribs)


 * Yes, it can be done, and fairly easily with a frontal shot and a profile. Of course, this will only give you half the face, since one profile only shows half the face.  You could then either mirror that half of the face to get the other side (which would only be an approximation, since faces aren't completely symmetrical), or use another profile shot to get the other side.  The 3D reconstruction could also be done using different angles, but would be more complex and the results less accurate.  In any case, the trickiest bit is mapping each point from one pic to the other, say the tip of the nose, corners of the mouth and eyes, etc. StuRat (talk) 17:32, 26 August 2011 (UTC)


 * As far as alternative methods, I suppose you could take a single frontal shot, and project it onto a "standard face". You'd need to pick a standard face from a library based on the face outline, eye position and size, mouth position and size, etc.  The height of the nose, etc., might not be quite right, though.  Note that facial hair would be most difficult to map using this method, but would also be a problem with any method. StuRat (talk) 17:40, 26 August 2011 (UTC)

Some research paper links: Generation of a 3-D Face Model from One Camera; Stereo photogrammetry for generating and matching facial models. We also have an article on photogrammetry. You might also want to look at geometric modeling. There are numerous ways to represent faces: feature-models; point clouds; polygon meshes; eigenface or generalized geometric basis coefficients; and so on. What are you looking for? Nimur (talk) 18:12, 26 August 2011 (UTC)

i think the problem is solved. i have exported .stl files using geomagic studios (from the wrl files in the database) and this is easily accessible through matlab. thanks anyway for your suggestions...--tathagata 18:27, 27 August 2011 (UTC) — Preceding unsigned comment added by Nonstop funstop (talk • contribs)