Test image 6

Blur kernel =        Noise level = 2% Mouse over the following links to flip between images By default, our result is shown MAP estimate - fixed sparse gradient prior, PSNR : 30.623607, SSIM : 0.957925 MAP estimate - adjusted sparse gradient prior, PSNR : 30.843703, SSIM : 0.958141 Degraded image Original image Gradient distribution User study result

Ours PSNR : 30.571094, SSIM : 0.957302

Mouse over the following links to flip between images By default, our result is shown MAP estimate - sparse gradient prior, PSNR : 30.623607, SSIM : 0.957925 MAP estimate - adjusted sparse gradient prior, PSNR : 30.843703, SSIM : 0.958141 Degraded image Original image Gradient distribution User study result

Test image 3

Blur kernel =        Noise level = 2% Mouse over the following links to flip between images By default, our result is shown MAP estimate - fixed sparse gradient prior, PSNR : 28.855561, SSIM : 0.961526 MAP estimate - adjusted sparse gradient prior, PSNR : 28.801853, SSIM : 0.958235 Degraded image Original image Gradient distribution User study result

Ours PSNR : 28.588971, SSIM : 0.958207

Mouse over the following links to flip between images By default, our result is shown MAP estimate - sparse gradient prior, PSNR : 28.855561, SSIM : 0.961526 MAP estimate - adjusted sparse gradient prior, PSNR : 28.801853, SSIM : 0.958235 Degraded image Original image Gradient distribution User study result