The effect of camera overhead on multi-image denoising performance:

The performance of multi-image denoising as the camera overhead varies (top curve). The horizontal lines at the bottom and in the middle indicate the PSNR of single image deblurring using a weak prior and a strong prior respectively. Multi-image denoising performs better than single-image deblurring, even when the camera overhead is 90%.

Multi-imag denoising result PSNR vs camera overhead

The following images show the results of multi-image denoising as the noise standard deviation increases with the camera overhead. These images correspond to the 10 data points on the curve in the above figure.

Denoising result for overhead 0.0, noise std = 20
Denoising result for overhead 0.1, noise std = 21.08
Denoising result for overhead 0.2, noise std = 22.36
Denoising result for overhead 0.3, noise std = 23.90
Denoising result for overhead 0.4, noise std = 25.82
Denoising result for overhead 0.5, noise std = 28.28
Denoising result for overhead 0.6, noise std = 31.62
Denoising result for overhead 0.7, noise std = 36.51
Denoising result for overhead 0.8, noise std = 44.72
Denoising result for overhead 0.9, noise std = 63.25