1: \begin{abstract}

2: For the constrained LiGME model, a nonconvexly regularized least squares estimation model, we present an iterative algorithm of guaranteed convergence to its globally optimal solution.

3: The proposed algorithm can deal with two different types of constraints simultaneously. 

4: The first type constraint, called the asymptotic one, requires the limit of estimation sequence to achieve the corresponding condition.

5: The second type constraint, called the early one, requires every vector in estimation sequence to achieve the condition. 

6: We also propose a bivariate nonconvex enhancement of fused lasso models with effective constraint

7: for sparse piecewise constant signal estimations.

8: (This is an improved version of [Yata and Yamada, ICASSP 2024].)

9: \end{abstract}

10: