Open aonihao1 opened 1 year ago
Hey, sorry for the late reply. Intents are forecasted using a different model, you can take the labels provided in the data folder. Also constant labels (eg: walking) can be used to predict the poses for waling action.
How do I use this network model? There are no details in the readme, such as how to train, how to get the intent tag, how to combine the intent tag with the action prediction, etc