a1302z / hierarchical_visual_localisation Goto Github PK

Where is the pretrained model and where should I put it?

Hi @a1302z Thanks for your excellent work!
How did you get the SFM database and Keypoint coordinates?
how to get these files: data/images_points.db, data/points3D.txt, data/images.txt if the local method I choose is colmap ？
I use colmap to get the sfm model, sparse reconstruction generate multiple output instead of one.
Looking forward to your guidance , thanks!

This following line results in RuntimeError: invalid argument 5: k not in range for dimension at /pytorch/aten/src/THC/generic/THCTensorTopK.cu:23 error

values, indices = torch.topk(d, n_neighbors, dim=1, largest=False, sorted=True)

the tensor size of d is (2000,1) with n_neighbors = 2. If I change n_neighbors to 1, it works but it's inaccurate. Any ideas?

I'm getting the following error while trying to run the evaluate.

  File "/home/Workspace/Python/hierarchical_visual_localisation/evaluate.py", line 1043, in <module>
    points3d, images, database_cursor, query_cursor, img_cluster, camera_matrices, query_images, query_image_ids, setup_time = setup(args)
  File "/home/Workspace/Python/hierarchical_visual_localisation/evaluate.py", line 443, in setup
    images = get_images()
  File "/home/Workspace/Python/hierarchical_visual_localisation/dataset_loaders/txt_to_db.py", line 44, in get_images
    for row in c.execute('SELECT image_id, qvec, tvec, camera_id, name, data_cols, xys, point3D_ids FROM images;'):
sqlite3.OperationalError: no such column: qvec

I have ofcourse downloaded the db file