Over the last decade, the demand for better segmentation and classification algorithms in 3D spaces has significantly grown due to the popularity of new 3D sensor technologies and advancements in the field of robotics. Point-clouds are one of the most popular representations to store a digital description of 3D shapes. However, point-clouds are stored in irregular and unordered structures, which limits the direct use of segmentation algorithms such as Convolutional Neural Networks. The objective of our work is twofold: First, we aim to provide a full analysis of the PointNet architecture to illustrate which features are being extracted from the point- clouds. Second, to propose a new network architecture called IPC-Net to improve the state-of-the-art point cloud architectures. We show that IPC-Net extracts a larger set of unique features allowing the model to produce more accurate segmentations compared to the PointNet architecture. In general, our approach outperforms PointNet on every family of 3D geometries on which the models were tested. A high generalisation improvement was observed on every 3D shape, especially on the rockets dataset. Our experiments demonstrate that our main contribution, inter-point activation on the network's layers, is essential to accurately segment 3D point-clouds.
Gomez Marulanda, F, Libin, P, Verstraeten, T & Nowe, A 2018, IPC-Net: 3D point-cloud segmentation using deep inter-point convolutional layers. in Proceedings - 2018 IEEE 30th International Conference on Tools with Artificial Intelligence, ICTAI 2018. vol. 2018-November, 8576052, Proceedings - International Conference on Tools with Artificial Intelligence, ICTAI, vol. 2018-November, IEEE, pp. 293-301, International Conference on Tools with Artificial Intelligence, Volos, Greece, 5/04/19. https://doi.org/10.1109/ICTAI.2018.00054
Gomez Marulanda, F., Libin, P., Verstraeten, T., & Nowe, A. (2018). IPC-Net: 3D point-cloud segmentation using deep inter-point convolutional layers. In Proceedings - 2018 IEEE 30th International Conference on Tools with Artificial Intelligence, ICTAI 2018 (Vol. 2018-November, pp. 293-301). Article 8576052 (Proceedings - International Conference on Tools with Artificial Intelligence, ICTAI; Vol. 2018-November). IEEE. https://doi.org/10.1109/ICTAI.2018.00054
@inproceedings{add4e3fc9fe448ab9411c4fe72eff79f,
title = "IPC-Net: 3D point-cloud segmentation using deep inter-point convolutional layers",
abstract = "Over the last decade, the demand for better segmentation and classification algorithms in 3D spaces has significantly grown due to the popularity of new 3D sensor technologies and advancements in the field of robotics. Point-clouds are one of the most popular representations to store a digital description of 3D shapes. However, point-clouds are stored in irregular and unordered structures, which limits the direct use of segmentation algorithms such as Convolutional Neural Networks. The objective of our work is twofold: First, we aim to provide a full analysis of the PointNet architecture to illustrate which features are being extracted from the point- clouds. Second, to propose a new network architecture called IPC-Net to improve the state-of-the-art point cloud architectures. We show that IPC-Net extracts a larger set of unique features allowing the model to produce more accurate segmentations compared to the PointNet architecture. In general, our approach outperforms PointNet on every family of 3D geometries on which the models were tested. A high generalisation improvement was observed on every 3D shape, especially on the rockets dataset. Our experiments demonstrate that our main contribution, inter-point activation on the network's layers, is essential to accurately segment 3D point-clouds.",
keywords = "Convolutional Neural Networks, Irregular Domains, Kernel visualisation, Segmentations",
author = "{Gomez Marulanda}, Felipe and Pieter Libin and Timothy Verstraeten and Ann Nowe",
year = "2018",
month = dec,
day = "13",
doi = "10.1109/ICTAI.2018.00054",
language = "English",
isbn = "978-1-5386-7449-9",
volume = "2018-November",
series = "Proceedings - International Conference on Tools with Artificial Intelligence, ICTAI",
publisher = "IEEE",
pages = "293--301",
booktitle = "Proceedings - 2018 IEEE 30th International Conference on Tools with Artificial Intelligence, ICTAI 2018",
note = "International Conference on Tools with Artificial Intelligence ; Conference date: 05-04-2019 Through 08-04-2019",
url = "http://ictai2018.org/",
}