Can the spherical harmonic coefficient difference method achieve good results when qin is 10 or above, better than RCNN

[wangshuangxing]

Why did I use the method of spherical harmonic coefficient difference to obtain better results than the PSNR and SSIM of RCNN

[m-lyon]

That may well be the case, the higher the q_in, the better a reconstruction spherical harmonics will give you. Likely it will depend though on what downstream analysis method you are comparing against.