1: \begin{abstract}

2:     It is well known that popular optimization techniques can lead to overfitting or even a lack of convergence altogether; thus, practitioners often utilize ad hoc regularization terms added to the energy functional.

3:     When carefully crafted, these regularizations can produce compelling results.

4:     However, regularization changes both the energy landscape and the solution to the optimization problem, which can result in underfitting.

5:     Surprisingly, many practitioners both add regularization and claim that their model lacks the expressivity to fit the data.

6:     Motivated by a geometric interpretation of the linearized search space, we propose an approach that ameliorates overfitting without the need for regularization terms that restrict the expressiveness of the underlying model.

7:     We illustrate the efficacy of our approach on minimization problems related to three-dimensional facial expression estimation where overfitting clouds semantic understanding and regularization may lead to underfitting that misses or misinterprets subtle expressions.

8: \end{abstract}

9: