D-DPCC: Deep Dynamic Point Cloud Compression via 3D Motion Prediction

Authors: Tingyu Fan, Linyao Gao, Yiling Xu, Zhu Li, Dong Wang

IJCAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental The experimental result shows that the proposed D-DPCC framework achieves an average 76% BD-Rate (Bjontegaard Delta Rate) gains against state-of-the-art Videobased Point Cloud Compression (V-PCC) v13 in inter mode.
Researcher Affiliation Collaboration Tingyu Fan 1 , Linyao Gao 1 , Yiling Xu1 , Zhu Li2 and Dong Wang3 1Cooperative Medianet Innovation Center, Shanghai Jiao Tong University 2University of Missouri, Kansas City 3Guangdong OPPO Mobile Telecommunications Corp., Ltd. {woshiyizhishapaozi, linyaog, yl.xu}@sjtu.edu.cn, zhu.li@ieee.org, wangdong7@oppo.com
Pseudocode No The paper does not contain structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide concrete access to source code for the methodology described.
Open Datasets Yes We train the proposed model using Owlii Dynamic Human DPC dataset [Keming et al., 2018], containing 4 sequences with 2400 frames... Following the MPEG common test condition (CTC), we evaluate the performance of the proposed D-DPCC framework using 8i Voxelized Full Bodies (8i VFB) [d Eon et al., 2017], containing 4 sequences with 1200 frames.
Dataset Splits No The paper mentions training on the Owlii dataset and evaluating on the 8i VFB dataset, but it does not specify a validation dataset or split for hyperparameter tuning.
Hardware Specification Yes We conduct all the experiments on a Ge Force RTX 3090 GPU with 24GB memory.
Software Dependencies No The paper mentions using an "Adam [Kingma and Ba, 2015] optimizer" but does not specify any software libraries with version numbers.
Experiment Setup Yes We train D-DPCC with λ =3, 4, 5, 7, 10 for each rate point. We utilize an Adam [Kingma and Ba, 2015] optimizer with β = (0.9, 0.999), together with a learning rate scheduler with a decay rate of 0.7 for every 15 epochs. A two-stage training strategy is applied for each rate point. Specifically, for the first five epochs, λ is set as 20 to accelerate the convergence of the point cloud reconstruction module; then, the model is trained for another 45 epochs with λ set to its original value. The batch size is 4 during training.