HM-PCGC: A HUMAN-MACHINE BALANCED POINT CLOUD GEOMETRY COMPRESSION SCHEME

Point cloud compression has various purposes in different application scenarios, such as requiring data fidelity in human vision tasks and pursuing semantic fidelity in machine vision tasks. This paper introduces a Human-Machine balanced point cloud geometry compression scheme (HM-PCGC) which considers the features of a variety of tasks. Our proposed scheme starts from a pre-trained, lightweight point cloud compression backbone and employs a Learned Semantic Mining module to aggregate multi-tasks features. By leveraging the aggregated features, HM-PCGC is able to retain the geometry and semantic properties of the point clouds, ensuring that they are accurately represented during compression. To better strikes a balance between the signal distortion and semantic distortion, we integrated a multi-task learning mechanism during the training phase. Our approach was extensively evaluated and analyzed, and the results demonstrate that it outperforms traditional and deep learning based point cloud codecs for both signal reconstruction and machine vision tasks.