Null-Space Diffusion Sampling for Zero-Shot Point Cloud Completion

Authors: Xinhua Cheng, Nan Zhang, Jiwen Yu, Yinhuai Wang, Ge Li, Jian Zhang

IJCAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental 5 Experiments; 5.1 Setups Evaluation Metrics. We evaluate the completion performance by computing the Chamfer Distance (CD) between the predicted completion results and the ground truth point clouds following previous methods [Tchapmi et al., 2019; Pan et al., 2021]; 5.2 Point Cloud Completion on Shape Net Datasets.
Researcher Affiliation Academia Shenzhen Graduate School, Peking University, China {chengxinhua, zhangnan}@stu.pku.edu.cn, zhangjian.sz@pku.edu.cn
Pseudocode Yes Algorithm 1 Null-Space Diffusion Sampling.
Open Source Code No The paper does not provide concrete access to source code for the methodology described.
Open Datasets Yes We conduct point cloud completion experiments on three datasets derived from Shape Net [Chang et al., 2015] benchmark with varying forms of degradation. The first dataset is a widely used benchmark PCN [Yuan et al., 2018], which is the subset derived from the Shape Net dataset. The second dataset is provided by Gen Re [Zhang et al., 2018].
Dataset Splits No The paper mentions test sets for evaluation but does not provide specific training/validation/test dataset splits for reproduction of the full experimental setup, nor does it explicitly mention a validation set.
Hardware Specification No The paper does not provide specific hardware details used for running its experiments.
Software Dependencies No The paper mentions using 'pre-trained Point-Voxel Diffusion (PVD) [Zhou et al., 2021]' but does not provide specific version numbers for PVD or any other software dependencies.
Experiment Setup Yes For hyper-parameters assigned in the tolerant loop mechanism, we set the loop count N = 20, the loop start time-step L = 100, and the null-space sampling interval K = 10.