Holmes-Alan

you have check your image size and patch size you use for testing. The maximum patch size is 64

just run the testing code. It super-resolve the image patch by patch for reconstruction.

Thanks for the comments. The reason we use 8 augmented version of LR patch is that we try to average the SR results to obtain more smooth quality and higher...

You asked a really good question. The key issue is that the spatial attention is a correlation model that cost many computation. So far 64 is the trade-off between quality...

Thanks for your comment. For the inner structure of the back projection block, it is modified from our previous work HBPN (http://openaccess.thecvf.com/content_CVPRW_2019/html/NTIRE/Liu_Hierarchical_Back_Projection_Network_for_Image_Super-Resolution_CVPRW_2019_paper.html). We did not touch too much on the...

> If I understand correctly, maybe you can convert quaternions to axis-angle or rotation matrix, and then use the forward function of SMPL. How did you convert quaternions to axis-angle...

Thanks for the comment. I wonder if I am using it correctly. I try to use the following code to transfer rot_mat to pose as pred_pose = output.rot_mats.float().reshape(-1, 4) pred_pose...

Thanks!!

> Hi @Holmes-Alan - I also think this package is super interesting. > > I wonder if what you're describing is possible via the [`scene.export()`](https://nerfvis.readthedocs.io/en/latest/nerfvis.html#nerfvis.Scene.export) method. I think you would...

> > Hi @Holmes-Alan - I also think this package is super interesting. > > I wonder if what you're describing is possible via the [`scene.export()`](https://nerfvis.readthedocs.io/en/latest/nerfvis.html#nerfvis.Scene.export) method. I think you...