Clarification on the Purpose and Usage of Diffusion-Generated Images and imageQueue in sim_manager_vad.py

[Zhzdpaaa]

I would like to understand the purpose of calling WorldDreamer and using diffusion to generate images in TrafficManager/sim_manager_vad.py. Are these images intended for the simulator’s rendering, or are they inputs for the VAD model?

I noticed that the three images returned by diffusion and the BEV image output by the VAD are combined into a variable ci, which is then self.model.imageQueue.put(ci). However, I couldn’t find where imageQueue.get() is called afterward.

Could you kindly clarify this process? Thank you very much for your response!