Gaussian Mixture Model Plot for Soft Assignment by Color

Question

enter image description here)

I want to show soft assignment by probabilities from GMM looking similar to the image posted. I want to show a data points membership between two (or more clusters).

Any help would be appreciated.

I tried to multiply probabilities from maximum of predict_proba(X) but it seems not working as it is not applied to labels=0.

ax.scatter(X[:, 0], X[:, 1], c=labels*probability)