Relevant instructions on training and testing

[P1ng3609]

Are there instructions related to training and testing, please? I noticed that there is no relevant information in the readme. It would be great if you could provide.

[callsys]

The proposed data generation pipeline relies on off-the-shelf trained models and requires no training. For testing, we have released the data generation code here. To evaluate the generated data quality, you can generate some videos and use them to train existing video text spotters.