TR2021-066

Wireless 3D Point Cloud Delivery Using Deep Graph Neural Networks


    •  Fujihashi, T., Koike-Akino, T., Chen, S., Watanabe, T., "Wireless 3D Point Cloud Delivery Using Deep Graph Neural Networks", IEEE International Conference on Communications (ICC), June 2021.
      BibTeX TR2021-066 PDF
      • @inproceedings{Fujihashi2021jun,
      • author = {Fujihashi, Takuya and Koike-Akino, Toshiaki and Chen, Siheng and Watanabe, Takashi},
      • title = {Wireless 3D Point Cloud Delivery Using Deep Graph Neural Networks},
      • booktitle = {IEEE International Conference on Communications (ICC)},
      • year = 2021,
      • month = jun,
      • url = {https://www.merl.com/publications/TR2021-066}
      • }
  • MERL Contact:
  • Research Areas:

    Communications, Machine Learning, Signal Processing

Abstract:

In typical point cloud delivery, a sender uses octreebased and graph-based digital video compression to send three dimensional (3D) points and color attributes. However, the digital-based schemes have an issue called the cliff effect, where the 3D reconstruction quality will be a step function in terms of wireless channel quality. To prevent the cliff effect subject to channel quality fluctuation, we have proposed a wireless point cloud delivery called HoloCast inspired by soft delivery. Although the HoloCast realizes graceful quality improvement according to instantaneous wireless channel quality, it requires large communication overheads. In this paper, we propose a novel scheme for soft point cloud delivery to simultaneously realize better 3D reconstruction quality and lower communication overheads. The proposed scheme introduces an end-to-end deep learning framework based on graph neural network (GNN) to reconstruct high-quality point clouds from its distorted observation under wireless fading channels. We demonstrate that the proposed GNN-based scheme can reconstruct a clean 3D point cloud with low overheads by removing fading and noise effects.