VFA
===

Overview
--------
Vector Field Attention (VFA) models in MIR provide attention-guided feature
extraction for deformable registration. This page documents the VFA family,
including TransVFA and HyperVFA.

Included variants
-----------------
- VFA
- TransVFA (transformer-augmented VFA)
- HyperVFA (hypernetwork-conditioned VFA)

Key modules
-----------
- ``MIR.models.VFA``
- ``MIR.models.TransVFA``
- ``MIR.models.HyperVFA``
- ``MIR.models.configs_VFA``

Related tutorials
-----------------
- `LUMIR24 VFA <https://github.com/junyuchen245/MIR/tree/main/tutorials/lumir24_VFA>`_
- `LUMIR25 VFA <https://github.com/junyuchen245/MIR/tree/main/tutorials/lumir25_VFA>`_
- `VFA image segmentation <https://github.com/junyuchen245/MIR/tree/main/tutorials/VFA_image_seg>`_
- `VFA image synthesis <https://github.com/junyuchen245/MIR/tree/main/tutorials/VFA_image_synthesis>`_