how to inference on internet image?

autonomousvision / monosdf

[NeurIPS'22] MonoSDF: Exploring Monocular Geometric Cues for Neural Implicit Surface Reconstruction

MIT License

573 stars 53 forks source link

Closed lucasjinreal closed 2 years ago

lucasjinreal commented 2 years ago

Does it requires a additional camera intrinsics?

niujinshuchong commented 2 years ago

Yes, you need to have camera poses and intrinsics.

lucasjinreal commented 2 years ago

But how can I get a camera pose and intrinsics on internet images?

niujinshuchong commented 2 years ago

Maybe you could try Colmap?