The RD points on Kodak

[tangzhisen]

Thanks for sharing the code! If possible, could you provide the RD points (the value of x and y axis) on Kodak dataset? (Preferably at least 4 RD points.) I would like to compare these results with the RD performance of my algorithm.

[mx54039q]

As models are released, you can test them on Kodak dataset.
Here is the result tested on our device,
bpp_entroformer = np.array([0.0866,0.1452,0.2632,0.4058,0.5925,0.9313,1.2559,1.6738,1.9657])
psnr_entroformer = np.array([27.63,29.16,31.38,33.18,35.13,37.72,39.68,41.57,42.70])

[tangzhisen]

As models are released, you can test them on Kodak dataset. Here is the result tested on our device, bpp_entroformer = np.array([0.0866,0.1452,0.2632,0.4058,0.5925,0.9313,1.2559,1.6738,1.9657]) psnr_entroformer = np.array([27.63,29.16,31.38,33.18,35.13,37.72,39.68,41.57,42.70])

Thank you for your reply！However, the pre-trained models you released are optimized by MSE, can you provide rate-distortion points optimized by ms-ssim？That is, the rate-distortion results of Figure 12 in your paper.

[mx54039q]

As models are released, you can test them on Kodak dataset. Here is the result tested on our device, bpp_entroformer = np.array([0.0866,0.1452,0.2632,0.4058,0.5925,0.9313,1.2559,1.6738,1.9657]) psnr_entroformer = np.array([27.63,29.16,31.38,33.18,35.13,37.72,39.68,41.57,42.70])

Thank you for your reply！However, the pre-trained models you released are optimized by MSE, can you provide rate-distortion points optimized by ms-ssim？That is, the rate-distortion results of Figure 12 in your paper.

The results optimized by MS-SSIM,
bpp_entroformer = np.array([0.1017,0.1572, 0.1975, 0.2288, 0.3286, 0.4701, 0.6686,0.9211,1.2701,1.7122, 2.0283])
msssim_entroformer = np.array([0.94454,0.96267,0.97046,0.97500,0.98306,0.98875,0.99265,0.99522,0.99694,0.99807,0.99854])

[TNTWEN]

@mx54039q 你好，请问可以提供一下论文中parallel模型的PSNR和MS-SSIM的RD points 数据嘛，您提供的parallel模型只有四个而且是MSE优化的。 感谢感谢！