vismatch.im_models.roma

Classes

RomaMatcher([device, max_num_keypoints])
TinyRomaMatcher([device, max_num_keypoints])

class vismatch.im_models.roma.RomaMatcher(device='cpu', max_num_keypoints=2048, *args, **kwargs)

Bases: BaseMatcher

Parameters:

device (str)

dino_patch_size = 14

coarse_ratio = 0.6481481481481481

compute_padding(img0, img1)

preprocess(img, pad=False)

Parameters:

img (Tensor)

Return type:

<module ‘PIL.Image’ from ‘/home/docs/checkouts/readthedocs.org/user_builds/vismatch/envs/stable/lib/python3.11/site-packages/PIL/Image.py’>

class vismatch.im_models.roma.TinyRomaMatcher(device='cpu', max_num_keypoints=2048, *args, **kwargs)

Bases: BaseMatcher

Parameters:

device (str)

preprocess(img)