Much faster

• Core rewritten to use quantized lookup table for coefficient
• Support SSE2, SSE3, AVX2 and FMA3.
• Basically, MUCH faster.

JincResize v0.1.1

Now compiled with ICC14.

More optimized code, should run more than 2 times faster.

Jinc Resizer v0.1

optimize AVX some more, but still slower than SSE with MSVC