JonasFrey96

Exporting Voxels

1

Hello, Thanks a lot for your work. Is there an easy way to extract individual class probabilites of the semantic voxels and store them as paramters to generated mesh. Also...

Official ROS1 Support Request

10

Hello Nvidia, thanks a lot for your awesome work on Nvblox. It would be great if you could also provide an official ROS1 wrapper for this package. Most deployed robots...

Added seaborn "rdylbu" colormap. Tested with ANYmal simulation. 

Parsing a List of dataclasses

8

Hello, it would be great if we could add support to pass a List of dataclasses: ```python @dataclass class ExperimentParams(Serializable): @dataclass class LayerParams: name: str loss_scale: float = 1.0 target_layers:...

[FEA] CUDA accelerated foolfill algorithm with distance return

3

Hello everyone, Thanks for the well-maintained and fast image processing library. **Feature Description** In robotics, we often would like to solve some planning problem on cost maps - e.g. solving...

We still changing the batch size during learning given the change in the number of samples per forward pass.

Switching from dino to stego requires different hyper parameter

1

When switching from: seg -> stego - random feature -> stego - dino Completly different hyperparameters are needed for the anomaly detection: - confidence_std_factor (stego) 1 - and with dino...

Confidence score

4

 makes currently a lot of sense when we have a bimodal distribution but makes little sense when we don't have one

Objective: - Simplify code - Clearer interfaces and responsibilities of components - Support other geometric primitives for reprojection e.g. individual footholds - Support more supervision signals - e.g: binary, velocity...

We currently missed adding the legged robotics fork of ros_comm. https://github.com/leggedrobotics/ros_comm This should fix the delay bug that messages move back in time with size 0