Skip to content

The way to employ BA for fine-tuned VGGT #454

@Wangweizhen-S

Description

@Wangweizhen-S

Thanks for your excellent work. I‘m trying to finetuing VGGT on a certain dataset which lacks texuture. After getting a good initial pose and depth, I want to use BA to refine the result further. However, it will probably fails in the way using VGGSfm to extract corresponding points as the demo_colmap.py shows for the dataset's intrinsic properties. So, I wonder, how to use the track head in VGGT(I will fine tune on my dataset) to get the corresponding points beacuse the VGGT track head's input seemly are certain query points while the matched feature points are computed. Looking forward for your reply. Thanks!

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions