Differentiable Euler Characteristic Transforms for Shape Classification

DECT enables end-to-end learning of ECT, enhancing shape classification across various modalities.