GAMPoint: Geometry-augmented vision transformer with a multimodal-ready interface for 3D point-cloud segmentation

This repository does not redistribute the ScanNet, S3DIS, or any other original dataset files. Instead, it provides supplementary materials and information necessary to reproduce our experiments. The datasets used in our experiments were obtained through the Pointcept / Point Transformer V3 (PTv3) data preparation pipeline and associated documentation. We acknowledge the authors of Point Transformer V3 and the Pointcept project for providing the preprocessing framework and dataset preparation guidelines used in our research. Users who wish to reproduce our experiments should obtain the original datasets directly from their official sources and comply with the corresponding licenses and terms of use. We do not claim ownership of any original dataset content. Original dataset sources: ScanNet: https://www.scan-net.org/ S3DIS: http://buildingparser.stanford.edu/dataset.html Preprocessing reference: Pointcept / Point Transformer V3: https://github.com/Pointcept/Pointcept Dataset licenses: ScanNet: ScanNet Terms of Use S3DIS: Subject to the terms and conditions provided by the dataset owners This record is provided solely for research reproducibility and non-commercial academic purposes. Users are responsible for obtaining the original datasets from the official providers and ensuring compliance with all applicable licenses and usage restrictions.