Search papers, labs, and topics across Lattice.
This paper introduces AdamFlow, a novel optimization method for surface registration that generalizes the Adam optimizer to probability space for minimizing the sliced Wasserstein distance between surface meshes represented as probability measures. By leveraging the sliced Wasserstein distance, the method achieves log-linear computational complexity, enabling efficient registration. Experiments demonstrate AdamFlow's superior performance in both affine and non-rigid surface registration across diverse anatomical structures, along with theoretical analysis of its asymptotic convergence.
AdamFlow offers a computationally efficient and robust alternative to existing surface registration methods by reformulating the problem as distributional optimization with a generalized Adam optimizer in probability space.
Surface registration plays an important role for anatomical shape analysis in medical imaging. Existing surface registration methods often face a trade-off between efficiency and robustness. Local point matching methods are computationally efficient, but vulnerable to noise and initialisation. Methods designed for global point set alignment tend to incur a high computational cost. To address the challenge, here we present a fast surface registration method, which formulates surface meshes as probability measures and surface registration as a distributional optimisation problem. The discrepancy between two meshes is measured using an efficient sliced Wasserstein distance with log-linear computational complexity. We propose a novel optimisation method, AdamFlow, which generalises the well-known Adam optimisation method from the Euclidean space to the probability space for minimising the sliced Wasserstein distance. We theoretically analyse the asymptotic convergence of AdamFlow and empirically demonstrate its superior performance in both affine and non-rigid surface registration across various anatomical structures.
