Joint - -divergences reconstruction and non-convex sparse regularization for image clustering

Key Points

• Image clustering improves with joint alpha and beta divergence approaches, enhancing partitioning accuracy.
• Performance metrics show significant reductions in error rates up to 15% during clustering tasks.
• Analysis of joint divergences alongside sparse regularization reveals a non-convex optimization landscape.
• This work could lead to more robust clustering methods in image processing, requiring validation in real-world scenarios.