Hi,

I was trying to run this program using the BZ Protogen Kit (available at https://drive.google.com/drive/folders/1JPh85gOn74EI8hujZB8KZcrOkNDQmolt) models converted from Blender to VRM using the Blender VRM addon with no extra modification.

However when I selected the model, the program crashes whenever it tries to load it in. The following is the error message produced:

Traceback (most recent call last):
File "vtuber_webcam.pyx", line 263, in init vtuber_webcam
File "vtuber_webcam.pyx", line 119, in vtuber_webcam.App.__init__
File "gltf.pyx", line 88, in plugins.avatar.glb_read
File "gltf.pyx", line 88, in plugins.avatar.glb_read
File "/home/USERNAME/Documents/vtuber-webcam-linux-64/python39.zip/gltf/converter.py", line 397, in update
File "/home/USERNAME/Documents/vtuber-webcam-linux-64/python39.zip/gltf/converter.py", line 229, in add_node
File "/home/USERNAME/Documents/vtuber-webcam-linux-64/python39.zip/gltf/converter.py", line 1155, in build_character
File "/home/USERNAME/Documents/vtuber-webcam-linux-64/python39.zip/gltf/converter.py", line 1292, in build_character_joints
KeyError: 'inverseBindMatrices'

Possibly there is a problem with the model, but I have no idea.

Running into what may be a related error when loading a VRM immediately after it creates springbones:

Assertion failed: !is_empty() at line 1510 of panda/src/pgraph/nodePath.cxx
Traceback (most recent call last):
  File "vtuber_webcam.pyx", line 304, in vtuber_webcam.App._update
  File "mediapipe.pyx", line 734, in plugins.mediapipe.VPlugin.update
  File "mediapipe.pyx", line 357, in plugins.mediapipe.VPlugin.reset
  File "cc.pyx", line 100, in plugins.avatar.CharacterController.get_bone_pos
AssertionError: !is_empty() at line 1510 of panda/src/pgraph/nodePath.cxx
:task(error): Exception occurred in PythonTask _update
Traceback (most recent call last):
  File "vtuber_webcam.pyx", line 339, in init vtuber_webcam
  File "/home/makyo/vtuber-webcam/python310.zip/direct/showbase/ShowBase.py", line 3328, in run
  File "/home/makyo/vtuber-webcam/python310.zip/direct/task/Task.py", line 546, in run
  File "/home/makyo/vtuber-webcam/python310.zip/direct/task/Task.py", line 497, in step
  File "vtuber_webcam.pyx", line 304, in vtuber_webcam.App._update
  File "mediapipe.pyx", line 734, in plugins.mediapipe.VPlugin.update
  File "mediapipe.pyx", line 357, in plugins.mediapipe.VPlugin.reset
  File "cc.pyx", line 100, in plugins.avatar.CharacterController.get_bone_pos
AssertionError: !is_empty() at line 1510 of panda/src/pgraph/nodePath.cxx

Same seems to happen to me, after loading spring bones the app crashes

chain "Tail.001" with 0 nodes was used to create a spring with 0 segments
:pnmtext: Loaded font Perspective Sans Regular
marionette thread started
./vtuber_webcam.sh: line 6: 87519 Segmentation fault      (core dumped) ./vtuber_webcam ${*}

I have been waiting for linux native app for this so if there's something i can do to help further please let me know!
Edit: It could be the model as well but VSeeFace seems to load the model just fine

I have updated the app. I hope it fixes those bugs.