I didn't know where I was going with this image, but wanted to try some new ideas.
Approximate work flow:
Reduced red eye and darkened pupils.
Extracted figure to be placed on a new background.
New background created from colors in original picture to help mask the extraction.
Cleaned up extraction edges with a layer mask and the blur tool.
Wall texture on it's own Soft Light layer for adjusting opacity and blurring.
Floor created from PamR's ideas about using vertical motion blur on a noise layer, then perspective transform. Blur tool used with receding distance.
Placed a merged visible layer at the bottom of the stack, blurred and stretched into a new enlarged canvas.
Stroked the image 1 pixel black, then added a drop shadow for the second offset line.
Drop shadow of the figure placed on its own layer and warped shape then gradient layer mask.
(In hind site, I should have had the shadow fall to the left, as in the original photo.)
I added a bevel/emboss to the cutout figure to see if that would soften the cut-out and give it a little 3-D modeling.
Layer 3 is dark vignette.
Layer 2 has no function (I forgot to delete it from an earlier experiment).




forum discussion: http://forums.dpreview.com/forums/read.asp?forum=1006&message=25968069

original image by Shaolin101: