Antti Isosalo
Antti Isosalo
Would it be possible to have also grey-scale ImageNet weights for the usual models along the lines described in Xie & Richmond? > Xie, Y. and Richmond, D., “[Pre-training on...
How about https://github.com/uoip/transforms/blob/master/transforms.py#L1004-L1054
Keypoints could be given as an image :) of some type and then after the transformation extracted from the distorted image?
> Example with satellite images For example something like reported in [this](https://www.isprs-ann-photogramm-remote-sens-spatial-inf-sci.net/IV-2-W7/47/2019/isprs-annals-IV-2-W7-47-2019.pdf) paper (Data Augmentation Approaches for Satellite Image Super-Resolution)?
```yaml {% set name = "solt" %} {% set version = "0.1.9" %} package: name: {{ name|lower }} version: {{ version }} source: url: https://pypi.io/packages/source/{{ name[0] }}/{{ name }}/solt-{{ version...
I suppose better would be to build (EDIT: the recipe) fully on `conda` and remove references to `pip` if possible? At least all the listed requirements seem to be available...
Would you be willing to update your repository with the added functionality in the meantime?
Tips from https://github.com/pytorch/pytorch/issues/8637 might come handy when adjusting the method