Is there an implementation of PyTorch to work with 3d object comparison, specifically with scan data? For instance, training a model on initial scans and corrected versions. For context I’m an artist with no development experience, but it seems PyTorch may have some capability to do this.