My first LoRA, trained to depict tiny flattened vehicles on a character's soles. Training data made of 100% furries, so I'm not confident in it working with humans. The cars can look a bit too realistic at times, as a byproduct of the way the training images were made.

Trigger Word: carsticker

I've also seen that adding car-related things into the positive prompt helps. Examples include "cars stuck to sole".