Johan Edstedt

### Describe the bug warp_perspective allows the ``align_corners = False`` argument, but only applies it to F.grid_sample and not to the meshgrid. As a result, the coordinate systems are inconsistent,...

Assumed coordinate systems for draw_rectangle and draw_convex_polygons are different

9

### Describe the bug Draw rectangle assumes (by casting coordinates to int), a coordinate system with the center of the top-left pixel at [0.5,0.5], while draw_convex_polygons assumes [0,0] at the...

Provide additional details on HPatches evaluation

8

Hi, I'm interested in how you calculated the AUC for homographies on HPatches exactly. Related: [#105](https://github.com/zju3dv/LoFTR/issues/105#issue-1051619274)

Depth images not aligning well

8

Hi, I'm trying to get the depth to align between images. However, my results typically look something like this:   When swapping between the images it is quite clear...

This solves https://github.com/chenxi116/DeepLabv3.pytorch/issues/14

Cov matrix used instead of precision matrix?

3

At: https://github.com/izmailovpavel/flowgmm/blob/422ff5dc15c5d2df6ada4787994d5915e11fcd4b/flow_ssl/distributions.py#L22-L25 You seem to insert a precision matrix, but reading the documentation: https://github.com/pytorch/pytorch/blob/8b248af35d43c97d0e437f6f4ff0fbd4da5700c8/torch/distributions/multivariate_normal.py#L119 It seems like if you do not specify the matrix as a precision matrix, it will...

Future Feature Request: Vectorization

1

Pytorch is (slowly) introducing vmap: https://pytorch.org/docs/master/generated/torch.vmap.html?highlight=vmap#torch.vmap When this feature becomes stable it seems like a great addition for jacobian calculation, probably giving additional performance.

WIP: Add DeDoDe detector+descriptor

2

#### Changes Adds dedode: https://github.com/parskatt/dedode #### Type of change - [ ] 📚 Documentation Update - [ ] 🧪 Tests Cases - [ ] 🐞 Bug fix (non-breaking change which...

Optimal Triangulation

2

## 🚀 Feature Currently kornia supports linear triangulation, adding optimal triangulation would make kornia more viable for triangulation of points from correspondences. ## Motivation Optimal triangulation is better than non-optimal...

Support float/double image width and height

1

Hi, Currently the python wrapper, see e.g. here: https://github.com/danini/graph-cut-ransac/blob/37f7469d49dbedddeaa5a2c1427b70121375c0bd/src/pygcransac/src/gcransac_python.cpp#L721-L731 only supports integer image sizes. As far as I can tell, these are always cast to doubles, so it would make...