Take a photo of the baby on a carpet, if at all possible.
If you decide you can't re-take the photo of the baby, reconsider the compositing choices being made here.
Think about how close you need to be to a carpet to see that much detail. If I'm using the carpet as a reference for the size of this baby, I would say it's smaller than the size of my hand.
Also be aware that the baby's head is turned and the mask graphic does not reflect this. The nose ridge on the mask should be distorted to fit the baby's face; a grid warp tool (GIMP / Paint.NET) may be able to do that.
The back of the baby seems to be darker than the baby's torso, but the lighting on the carpet doesn't match. Likewise, the shading on the mask is much too intense for the light diffuse shading on the baby, or the even-toned shading of the carpet. Consider selecting props that match the lighting on the baby.
It looks like the picture of the baby was taken with a white light or indirect sunlight. The carpet picture appears to be under a yellowish light, and the mask is under an incredibly intense light. The intensities (and colors) of all these light sources should be homogeneous (or close enough to stave off suspension of disbelief).
One of the most important cues for proximity is shading. There aren't any discernible shadows around the baby that tell an observer that this baby is physically touching the carpet. Faking these shadows believably is a difficult endeavor because it must conform to the bumpy surface of the carpet. I implore you to re-take the photo of the baby.
The edges of the mask seems much sharper than the baby's face, so it appears that the mask is 'floating' in front of the baby's face.
It may be beneficial to look up compositing or the aspects of design associated with photography to gain a better understanding of what people think and perceive when viewing art.