voxelmorph and TransMorph_Transformer_for_Medical_Image_Registration
VoxelMorph is a foundational unsupervised learning framework for deformable image registration, while TransMorph builds upon and extends that approach by incorporating transformer architectures for improved registration performance, making them complementary methodologies rather than direct competitors.
About voxelmorph
voxelmorph/voxelmorph
Unsupervised Learning for Image Registration
Implements diffeomorphic deformation learning through velocity field integration and spatial transformer networks, supporting N-dimensional affine and dense transforms with multiple interpolation strategies. Provides SynthMorph for data-free training via synthetic image generation and HyperMorph for automatic hyperparameter tuning, eliminating manual registration configuration. Built on PyTorch (with stable TensorFlow legacy support) and handles medical imaging formats (NIfTI, MGZ, npz) with optional semi-supervised learning via segmentation labels.
About TransMorph_Transformer_for_Medical_Image_Registration
junyuchen245/TransMorph_Transformer_for_Medical_Image_Registration
TransMorph: Transformer for Unsupervised Medical Image Registration (PyTorch)
Scores updated daily from GitHub, PyPI, and npm data. How scores work