In the post process, I've done a bit of correcting on the hair and the shadow pass.
The facial hair were rendered as white on a black head, and the result you see were used as a mask on white layer, because you can then control transparency of those facial hairs if they look too strong. Same goes for the shadow pass layer.
For the background, I've just used a photo of my room. Actually there were a lot of background tests, but in the end this was most suitable to me.
In the end, I composed the background with the final model, made some post process like colour correction, leveling and etc. and then applied depth of field using a zdepth map.