Download code/tt_superpoint/__init__.py from changh95/superpoint-p150: direct link, hf CLI and curl.
- Browser
- Download file 725 Bytes
-
https://huggingface.co/changh95/superpoint-p150/resolve/main/code/tt_superpoint/__init__.py
- Command line
-
hf download hf://changh95/superpoint-p150/code/tt_superpoint/__init__.py
-
curl -L -o __init__.py https://huggingface.co/changh95/superpoint-p150/resolve/main/code/tt_superpoint/__init__.py
725 Bytes
| # SPDX-License-Identifier: Apache-2.0 | |
| """SuperPoint keypoint detection + description on a Tenstorrent Blackhole chip (tt-nn). | |
| from tt_superpoint import SuperPoint | |
| with SuperPoint.from_pretrained(device_id=0) as model: | |
| out = model("image.jpg") # or a PIL image / numpy array / torch tensor | |
| print(out.keypoints.shape, out.scores[:5], out.descriptors.shape) | |
| See PYTHON.md for the full API reference. | |
| """ | |
| from .inputs import to_plane | |
| from .model import DEFAULT_MODEL_ID, DEFAULT_REVISION, INPUT_SIZE, SuperPoint, SuperPointOutput | |
| __version__ = "0.1.0" | |
| __all__ = ["SuperPoint", "SuperPointOutput", "to_plane", "DEFAULT_MODEL_ID", "DEFAULT_REVISION", "INPUT_SIZE", | |
| "__version__"] | |