Search results

Using panoramic images to train the VGGT model, the predicted depth map has all values being a constant distribution with very low variance.

Status: Open.

#475 In facebookresearch/vggt;

· Tony-Hou opened on May 2, 2026

Long video support: factor graph stitching pipeline (open source)

Status: Open.

#474 In facebookresearch/vggt;

· jashshah999 opened on May 2, 2026

The sequence of images will be divided into two scenes

Status: Open.

#472 In facebookresearch/vggt;

· xiazhenyu555 opened on Apr 3, 2026

Are VGGT output depth maps metric or relative?

Status: Open.

#471 In facebookresearch/vggt;

· sourcesync opened on Apr 2, 2026

I found that the cause of the OOM (Out of Memory) issue in demo_colmap is due to the resize dimension problem.

Status: Open.

#470 In facebookresearch/vggt;

· fee0103 opened on Apr 2, 2026

About DTU evaluation- aligning the two point clouds

Status: Open.

#469 In facebookresearch/vggt;

· MarthaCostella opened on Mar 15, 2026

Releasing evaluation code?

Status: Open.

#467 In facebookresearch/vggt;

· stephanie-fu opened on Mar 12, 2026

Visualized Network Attention Layer

Status: Open.

#466 In facebookresearch/vggt;

· gd9036 opened on Mar 9, 2026

CO3D evaluation preprocessing fails with single_sequence_subset

Status: Open.

#465 In facebookresearch/vggt;

· cavalier501 opened on Mar 9, 2026

How to evaluate the pose information output by VGGT

Status: Open.

#461 In facebookresearch/vggt;

· qxpBlog opened on Feb 28, 2026

CO3D annotation file camera mismatch

Status: Open.

#460 In facebookresearch/vggt;

· stephanie-fu opened on Feb 9, 2026

qvec issue always there when run the -use- ba

Status: Open.

#456 In facebookresearch/vggt;

· wuwei5v opened on Jan 5, 2026