vismatch.im_models

Wrapper modules for the supported matchers.

Importing vismatch.im_models.<wrapper> executes the wrapper inside its own ImportSandbox, so the third-party imports it performs (and the sys.path dirs it registers via add_to_path) stay private to that wrapper instead of leaking into the global interpreter state. See vismatch/import_sandbox.py for the full rationale.

Submodules

`vismatch.im_models.aff_steerers`
`vismatch.im_models.aspanformer`
`vismatch.im_models.dedode`
`vismatch.im_models.duster`
`vismatch.im_models.edm`
`vismatch.im_models.efficient_loftr`
`vismatch.im_models.gim`
`vismatch.im_models.handcrafted`
`vismatch.im_models.keypt2subpx`
`vismatch.im_models.kornia`
`vismatch.im_models.liftfeat`
`vismatch.im_models.lightglue`
`vismatch.im_models.lisrd`
`vismatch.im_models.loftr`
`vismatch.im_models.master`
`vismatch.im_models.matchanything`
`vismatch.im_models.matchformer`
`vismatch.im_models.matching_toolbox`
`vismatch.im_models.minima`
`vismatch.im_models.omniglue`
`vismatch.im_models.rdd`
`vismatch.im_models.ripe`
`vismatch.im_models.roma`
`vismatch.im_models.romav2`
`vismatch.im_models.se2loftr`
`vismatch.im_models.silk`	Standalone SiLK (Self-supervised Interest point Learning with Keypoints) implementation.
`vismatch.im_models.steerers`
`vismatch.im_models.topicfm`
`vismatch.im_models.ufm`
`vismatch.im_models.xfeat`
`vismatch.im_models.xfeat_steerers`
`vismatch.im_models.xoftr`
`vismatch.im_models.zippypoint`