Isleepinahammocks idea would probably work. But rotation and translation would not. Thats something you can easily take care of in your training data, by reusing the same training data in multiple random positions and random angles.
Yes i know. But this is trivially easy to get into the training data. You just slice models in random orientations and build your training dataset that way so the algortihm will recognize gun parts in every direction.
Isleepinahammocks idea would probably work. But rotation and translation would not. Thats something you can easily take care of in your training data, by reusing the same training data in multiple random positions and random angles.
are you familiar with 3d printing? The print head’s instructions will be completely different if you rotate the model. Unrecognizably different.
Yes i know. But this is trivially easy to get into the training data. You just slice models in random orientations and build your training dataset that way so the algortihm will recognize gun parts in every direction.