Tianyi Lim

Assuming you have a CUDA GPU on your system (like the Nvidia Jetson), you could try installing `nvcc` by running ``` sudo apt install nvidia-cuda-toolkit ```

Thanks for doing this!

If you're still stuck on this, take a look [here](https://arnon.dk/matching-sm-architectures-arch-and-gencode-for-various-nvidia-cards/). `sm_20` refers to a (very) old NVIDIA GPU architecture. If you're using a Jetson TX2, you likely need to change...

Hi, sorry for commenting on a closed issue. But if we are using the torch hub model, how do we modify it to use float? ```python3 # where to modify...

for the record, I found access to a computer with a newer GPU. Sorry :(

hi, if you are using an iphone, perhaps you need to modify the IMU data. You should flip the sign of ax ay and az and multiply them by 9.81...

You need to edit the BackendParams.yaml file. This should be found in `~/catkin_ws/src/Kimera-VIO/params/Euroc/BackendParams.yaml` Set `autoInitialize: 1`. Then you don't need to provide a GT Odom source for initialization.

Hi, were you able to resolve this?

Aside from the solutions in [this comment](https://github.com/UZ-SLAMLab/ORB_SLAM3/issues/443#issuecomment-1154814254) in #443, I found a potential other source of the bug. In [`src/System.cc`](https://github.com/UZ-SLAMLab/ORB_SLAM3/blob/4452a3c4ab75b1cde34e5505a36ec3f9edcdc4c4/src/System.cc#L1411-L1413): ```cpp string pathSaveFileName = "./"; pathSaveFileName = pathSaveFileName.append(mStrSaveAtlasToFile); pathSaveFileName =...

As an initial strategy, I wanted to reset the filter state, and use the last-known state as a initialization point. Looking at #398, re-starting the ROS node is not an...