Hugues THOMAS
Hugues THOMAS
Look at: https://github.com/HuguesTHOMAS/KPConv-PyTorch/blob/680296878d238e6bdb798c190120062a46f492d1/datasets/common.py#L34-L36 where the built function are imported and used in python
The validation split can only be an index. However, you can define multiple clouds to be in the same split For example, lets say your dataset is like this: ```...
As I said, you can define multiple clouds in the same split `self.all_splits = [0,1,0,0,0,1,0,0,0,0,1,0,0,0,0,1,0,0,0,0,1,0,0,0,0]` `self.validation_split = 1`
Can you make sure your feature matrix `batch.features` has 2 dimension (N, D) where N is the number of points and D the dimension of the features. In your case...
Maybe you need to recompile the cpp wrappers?
Hi @Pippo809, First of all you will need to change the loss of the network here: https://github.com/HuguesTHOMAS/KPConv-PyTorch/blob/1defcd75cf7c0399704a6a9f63d3a550bfb8c1c9/models/architectures.py#L151-L171 You can adapt this function for a regression task. I am not aware...
Hi, I am having the same issue here. I tried a bunch of different things to solve this without success. Here: https://github.com/ros2/ros1_bridge/blob/859e156aa351701442cdfe5507621b5bda86d1c8/CMakeLists.txt#L60-L61 I used `message()` to verify the path found...
I eventually solved this issue by changing the order of the source commands. the README says: ``` # You have already sourced your ROS installation. # Source your ROS 2...
If I understand correctly, you want to concatenate the features from all layers as a final feature, instead of only using the N1 x 64 last feature, right? In that...
No I have not