python script for general action recognition

[Ishihara-Masabumi]

I run iva.py and get the output for spin or squat.
But, I'd like to run the general action recognition.
Then, which python script should I use for the general action recognition?

[DEEPIKA-217M]

it looks like iva.py is currently acting as a demo script and may be limiting or simplifying how predictions are displayed.

Could you please confirm if iva.py is intended to:

Only visualize a subset of actions, or
Map full model outputs using a label file (e.g., labels.txt / labels.json)?

If the model already supports general action recognition, then the main improvement might be:

Ensuring full label mapping is loaded dynamically instead of hardcoded filtering
Updating output decoding so all supported classes are shown consistently

I’d be happy to help refactor the script to properly expose all predicted actions in a general inference mode.