Generate Images for Image Recognition

Hi, for my bachelor thesis in mathematics I wanted to train a neural network to classify street signs using images rendered in unreal engine and see how good this will then classify real world images. My question is whether or not I can use the following assets to generate those images in unreal engine:

(1) Can I use assets from quixel megascans?

(2) Can I use assets from the starter content/ assets from the learning tab like from the infiltrator demo?

(3) Can I use assets from the marketplace?

The end result will be images, so no compiled version of unreal engine and those images would then be used to train the neural network.

Megascans, starter content, learning tab assets, and marketplace assets are all intended for use in UE projects, including commercial products. It sounds like your neural network itself will not rely on UE, only the rendered images that you will use to train the network. If that’s the case, you should be clear to use all of the items you listed under the UE EULA.

Yes the neural network itself won’t rely on UE, only the rendered images. Thanks for your time answering this!