[Bug] Use mmaction2 - main/configs/skeleton/posec3d/rgbpose_conv3d/rgbpose_conv3d py error for the demo

[aonihao1]

Branch

main branch (1.x version, such as v1.0.0, or dev-1.x branch)

Prerequisite

• [X] I have searched Issues and Discussions but cannot get the expected help.
• [X] I have read the documentation but cannot get the expected help.
• [X] The bug has not been fixed in the latest version.

Environment

Python: 3.9.13 
GPU 0: NVIDIA GeForce GTX 1080 Ti
PyTorch: 1.12.1

Describe the bug

Traceback (most recent call last): File "F:\shiyan\mmaction2-main\demo\demo_skeleton.py", line 169, in main() File "F:\shiyan\mmaction2-main\demo\demo_skeleton.py", line 157, in main result = inference_skeleton(model, pose_results, (h, w)) File "F:\shiyan\mmaction2-main\mmaction\apis\inference.py", line 168, in inference_skeleton return inference_recognizer(model, fake_anno, test_pipeline) File "F:\shiyan\mmaction2-main\mmaction\apis\inference.py", line 105, in inference_recognizer data = test_pipeline(data) File "C:\Users\YA\anaconda3\envs\pytorch\lib\site-packages\mmengine\dataset\base_dataset.py", line 58, in call data = t(data) File "C:\Users\YA\anaconda3\envs\pytorch\lib\site-packages\mmcv\transforms\base.py", line 12, in call return self.transform(results) File "F:\shiyan\mmaction2-main\mmaction\datasets\transforms\pose_transforms.py", line 1342, in transform results['filename'] = results['frame_dir'] + '.mp4' KeyError: 'frame_dir'

Reproduces the problem - code sample

No response

Reproduces the problem - command or script

No response

Reproduces the problem - error message

No response

Additional information

No response