Topdown

This Lora was made as an extension or specialisation of the overheadview Lora (link below). It has been trained on overhead and aerial view images in a more general sense of objects and settings (boats, cars, cityscapes, etc) in addition to people, but does not have the same capability for angled shots of humans and poses that overheadview does. This one is just for straight-down shots, while overheadview can do more varied angles and poses, but is not fantastic at the top-down scenery.

/>
I use topdown at 0.6, and overheadview at 0.2 strength in the linked images. Together, they work well.