Skip to main content

Table 2 The learned \(\{\beta _k\}\) for real natural images with different levels of noise

From: Optimal graph edge weights driven nlms with multi-layer residual compensation

  \(\beta _0\) \(\beta _1\) \(\beta _2\) \(\beta _3\) \(\beta _4\)
\(\sigma <5\) 1.0058 0.0002 0.0003 − 0.0003 0.0052
\(\sigma =10\) 0.9996 0.0003 − 0.0001 − 0.0003 0.0069
\(\sigma =20\) 0.9946 0.0001 0.0001 0.0001 0.0046
\(\sigma =30\) 0.9940 0.0000 0.0001 0.0001 0.0070
\(\sigma =40\) 0.9955 − 0.0004 − 0.0003 − 0.0008 0.0153
\(\sigma =50\) 0.9903 − 0.0012 − 0.0016 − 0.0016 0.0085