Closed mtli77 closed 3 years ago
Hi @Violetit If you run the script in the same folder with tf_custom_ops/
, the program should be able to find it.
Hi @Violetit If you run the script in the same folder with
tf_custom_ops/
, the program should be able to find it.
Do you mean to put python evaluate_kitti_our.py
under the D3Feat/tf_custom_ops/
to run?
You can run python repeatability/evaluate_kitti_our.py
under the D3Feat
folder.
$export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:(path to were libtensorflow_framework.so is)
sh compile_op.sh
You can try it.
Hi @XuyangBai I have completed training on the kitti dataset. After generating the descriptors and detection scores, I ran the script
python evaluate_kitti_our.py D3Feat [timestr of the model]
to calculated the keypoint repeatability, but an error occur:However, no errors were reported during the entire training process of the kitti dataset, which should prove that my compilation is correct and the file is in the path
D3Feat/tf_custom_ops/tf_neighbors.so
.How can I help the
python evaluate_kitti_our.py D3Feat [timestr of the model]
to find theD3Feat/tf_custom_ops/tf_neighbors.so
file? Could you please help me to solve this problem?