1. Combining the Power of Internal & External Denoising Inbar Mosseri The Weizmann Institute of Science, ISRAEL ICCP, 2013

2. Outline  Introduction  Background  Patch_psnr  Results

3. Internal Denoising  NLM  BM3D Denoising using other noisy patches within the same noisy image

4. External Denoising  EPLL  Sparse Denoising using external clean natural patches or a compact representation

5. a) Originalb) Noisy inputc) Internal NLMd) External NLMe) Combinining (c)&(d) Internal vs. External Denoising

6. Internal vs. External Patch Preference the higher the noise in the image, the stronger the preference for internal denoising

7. PatchSNR

8. patches with low PatchSNR (e.g., in smooth image regions) tend to prefer Internal denoising patches with high PatchSNR (edges, texture) tend to prefer External denoising

9. Overfitting the Noise-Mean the empirical mean/variance of the noise within an individual small patch is usually not zero/ σ 2

10. Fitting of the Noise Mean The denoising error grows linearly with the deviation from zero of the empirical noise-mean within the patch. In contrast, the denoising error is independent of the empirical noise variance within the patch.

11. These patches are dominated by noise. There are high correlation between a random noise patch n and its similar natural patch NN(n) Overfit the Noise Detail

12. These patches are dominated by noise. There are high correlation between a random noise patch n and its similar natural patch NN(n)

14. Estimate the PatchSNR But var(n) is also unknown and patch- dependent.