search

JingyunLiang / RVRT

Recurrent Video Restoration Transformer with Guided Deformable Attention (NeurlPS2022, official repository)
https://arxiv.org/abs/2206.02146
Other
354 stars 33 forks source link

Runtime error occured, when it is used denoising. Who could help with this? #22

Open Dawson0813 opened 1 year ago

Dawson0813 commented 1 year ago

Run with bellow commod: python main_test_rvrt.py --task 001_RVRT_videosr_bi_REDS_30frames --folder_lq testsets/your/own --tile 0 0 0 --tile_overlap 2 20 20

Error detail infomation: File "/opt/conda/envs/env_rvrt/lib/python3.10/site-packages/torch/nn/modules/conv.py", line 613, in forward return self._conv_forward(input, self.weight, self.bias) File "/opt/conda/envs/env_rvrt/lib/python3.10/site-packages/torch/nn/modules/conv.py", line 608, in _conv_forward return F.conv3d( RuntimeError: Given groups=1, weight of size [192, 4, 1, 3, 3], expected input[1, 3, 38, 256, 256] to have 4 channels, but got 3 channels instead

kulius7 commented 1 year ago

I have the same issue, even when using images from the DAVIS or Set8 dataset without a folder_gt argument. The same error occurs in VRT. Also tried it locally on Ubuntu 22.04.2 with the exact same error. Fixes like setting -sigma 1 or nonblind_denoising == True are not working. I'd love to try out the denoising.

Dawson0813 commented 1 year ago

I have the same issue, even when using images from the DAVIS or Set8 dataset without a folder_gt argument. The same error occurs in VRT. Also tried it locally on Ubuntu 22.04.2 with the exact same error. Fixes like setting -sigma 1 or nonblind_denoising == True are not working. I'd love to try out the denoising.

I sloved this issue from bellow answers, hope to help with you: https://github.com/JingyunLiang/RVRT/issues/11