Molecular polarity results from the vector sum of the individual bond dipoles. #""^(delta-)C-H^(delta+)# are only very slightly polar; there is LITTLE electronegativity difference. In any case the VECTOR sum of the individual bond dipoles in a tetrahedral molecule is ZERO. (It would also be zero in #SiF_4#, even tho the #Si-F# bonds are also polar).
And likewise for carbon dioxide. Each #""^(delta+)C=O^(delta-)# is bond is quite polar, quite charge-separated. However, becaue carbon dioxide is LINEAR, the vector sum of the bond dipoles is clearly zero. Do you see this? You might have to cross-fertilize this idea of vector addition from your studies in physics: vectors have magnitude and direction, and they sum geometrically.
But for #SiH_2F_2# (which I think is what you meant to represent), you have #2xxSi-H# dipoles, which are not really polar, and #2xx""^(delta+)Si-F^(delta-)# bonds, the which are substantially polar. The vector sum of these dipoles, and I would consult a text pdq, results in a net molecular dipole moment. Thus #SiH_2F_2# is a polar molecule.