-
1. https://arxiv.org/abs/1811.02834
2. https://arxiv.org/abs/1805.09114
Understand if any variants of FGW allow us to substitute fast matrix vector multiplication or tweak them to be reusable with…
-
Hi! Thank you for your great work in advance
I ran your code and found some errors because of the object type
Errors occurred in the code block below (DWKM_utils.py):
```python
# Calcu…
-
Thank you for the excellent repo. I am looking at the example of (Fused) GW Dictionary Learning : https://pythonot.github.io/auto_examples/gromov/plot_gromov_wasserstein_dictionary_learning.html#estim…
-
As discussed previously, we have a pretty flat API at the moment.
One proposal would be to split it based on the problems:
```yml
- linear
- initializer.py
- problem.py
- solver.py
…
-
## Describe the bug
### To Reproduce
Steps to reproduce the behavior:
1. Calculating the barycenter with the optional arg `log=True`.
2. `log=True` in `gromov_wasserstein` returns an additiona…
-
Hello,
first of all, thank you very much for your repository. It was very useful for our project in which we need a fast method to calculate an approximation of the wasserstein barycenters.
But …
-
Dear @jeanfeydy ,
Thanks for your great. When will be the next update? e.g., out-of-the-box ImageLoss and VolumeLoss layers
![image](https://user-images.githubusercontent.com/19947331/83827364-c…
-
Hello,
If my understanding is correct, we should be able to use the Bottleneck distance for the TTK Persistance Diagram Clustering filter by setting the p parameter to "inf".
It is unfortunately n…
-
In paper
**_How to choose_** 
**Repository:** https://github.com/joliencremers/waspr
**Version:** v1.0.0
**Editor:** Pending
**Reviewer:** Pending
**Managing EiC:** Kristen Th…