Hello,
First of all, thanks for your very good work and dataset!
I would like to evaluate your algorithm using the full resolution images (1x).
The provided config file (humanrf/configs/example_humanrf.py) assumes quarter resolution images (4x) based on the --dataset.scale option.
Is there anything else I need to modify in order to achieve the best possible results at full resolution?
If so, could you please provide a config file covering this case?
Thanks :)
Hello, First of all, thanks for your very good work and dataset! I would like to evaluate your algorithm using the full resolution images (1x). The provided config file (humanrf/configs/example_humanrf.py) assumes quarter resolution images (4x) based on the --dataset.scale option. Is there anything else I need to modify in order to achieve the best possible results at full resolution? If so, could you please provide a config file covering this case? Thanks :)