It seems that the object 'reconstruction' is the wrong type before enter into features.FeatureView(), I also output feature_set
and reconstruction, they are:
<pixsfm._pixsfm._features.FeatureSet_f16 object at 0x7f876e078870>
<Reconstruction 'num_reg_images=10, num_cameras=10, num_points3D=1852, num_observations=8071'>
By the way, I can run the raw reconstruction without refinement. Please help me with the problem, thank you.
Hi. Thanks for your work. I met some problems running the Demo at:
refined, sfm_outputs = sfm.reconstruction(ref_dir, images, sfm_pairs, features, matches, image_list=references)
The output is: File /pixel-perfect-sfm/pixsfm/bundle_adjustment/main.py:133, in FeatureReferenceBundleAdjuster.refine(self, reconstruction, feature_set, problem_setup) 131 print(reconstruction) 132 # --> 133 feature_view = features.FeatureView( 134 feature_set, 135 reconstruction 136 ) 138 # Schedule reference computation (label for each point3D) 139 problem_labels = find_problem_labels(reconstruction, 140 self.conf.max_tracks_per_problem)TypeError: FeatureView(): incompatible function arguments. The following argument types are supported:
Invoked with: ,
It seems that the object 'reconstruction' is the wrong type before enter into features.FeatureView(), I also output feature_set and reconstruction, they are: <pixsfm._pixsfm._features.FeatureSet_f16 object at 0x7f876e078870> <Reconstruction 'num_reg_images=10, num_cameras=10, num_points3D=1852, num_observations=8071'>
By the way, I can run the raw reconstruction without refinement. Please help me with the problem, thank you.