You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Theory (Theorem 3 in this paper ) tells us that the Sinkhorn barycenter between two Gaussian distribution with the same std $\sigma$ should be a Gaussian with std $\sigma$.
However, when computing the barycenter between two Dirac-ish measures (Eulerian representation : measures are supported on a grid, with the mass concentrated on a single pixel), the barycentric interpolation using convolutional_barycenter2d_debiased returns a quite spread measure, which seems quite surprising.
To Reproduce
Steps to reproduce the behavior:
Define two measures on a grid of size (say) $20 \times 20$.
Compute there barycenter using convolutional_barycenter2d_debiased with parameter $t \in (0,1)$
Plot the result : it is quite blurry.
Screenshots
The "interpolation" we get by computing the Sinkhorn debiaised barycenter as a minimizer of $\mu \mapsto (1-t) S_\epsilon(\mu, \mu_0) + t S_\epsilon(\mu, \mu_1)$, with $\mu_0, \mu_1$ being two Dirac/very sharp Gaussian measures. The regularization parameter $\epsilon$ is set to reg=0.1.
Note : a similar behavior occurs when considering 1D measures, suggesting that the issue is not intrinsic to the convolutional-2D approach (did not investigated this in detail). The following screenshot show the returned barycenter for two 1D-Gaussian distributions, with reg=0.1 . The Gaussian distribution midway is too wide.
The barycenter of two Dirac-ish measures should be (close to, up to numerical approximation) a Dirac-ish measure supported midway. More generally, the barycenter of two gaussian with std $\sigma$ should have std $\sigma$.
Environment (please complete the following information):
Has been discussed with @hichamjanati (who agrees on the issue). We plan to investigate this later. This issue is for documentation---discussion is welcome!
The text was updated successfully, but these errors were encountered:
very good point, I was a bit underwhelemd by the debiased barycenters that did not look that much bebiased in practice in practice, maybe it is an implementation problem?
Yes it really looks like an implementation issue (or maybe something deeper, like the IBP not converging?).
I have no clue on what is happening yet. I did not investigate in details; I am quite busy these weeks (so is @hichamjanati ), but we may look at it closer later. In the meantime, if people have suggestions/remarks, there are welcome.
Describe the bug
Theory (Theorem 3 in this paper ) tells us that the Sinkhorn barycenter between two Gaussian distribution with the same std$\sigma$ should be a Gaussian with std $\sigma$ .
However, when computing the barycenter between two Dirac-ish measures (Eulerian representation : measures are supported on a grid, with the mass concentrated on a single pixel), the barycentric interpolation using
convolutional_barycenter2d_debiased
returns a quite spread measure, which seems quite surprising.To Reproduce
Steps to reproduce the behavior:
convolutional_barycenter2d_debiased
with parameterScreenshots
The "interpolation" we get by computing the Sinkhorn debiaised barycenter as a minimizer of$\mu \mapsto (1-t) S_\epsilon(\mu, \mu_0) + t S_\epsilon(\mu, \mu_1)$ , with $\mu_0, \mu_1$ being two Dirac/very sharp Gaussian measures. The regularization parameter $\epsilon$ is set to
reg=0.1
.Note : a similar behavior occurs when considering 1D measures, suggesting that the issue is not intrinsic to the convolutional-2D approach (did not investigated this in detail). The following screenshot show the returned barycenter for two 1D-Gaussian distributions, with
reg=0.1
. The Gaussian distribution midway is too wide.Code sample
To reproduce the plot in 2D :
Expected behavior
The barycenter of two Dirac-ish measures should be (close to, up to numerical approximation) a Dirac-ish measure supported midway. More generally, the barycenter of two gaussian with std$\sigma$ should have std $\sigma$ .
Environment (please complete the following information):
Additional context
Has been discussed with @hichamjanati (who agrees on the issue). We plan to investigate this later. This issue is for documentation---discussion is welcome!
The text was updated successfully, but these errors were encountered: