Point Set Voting for Partial Point Cloud Analysis

Zhang, Junming; Chen, Weijia; Wang, Yuping; Vasudevan, Ram; Johnson-Roberson, Matthew

doi:10.1109/lra.2020.3048658

Cited by 31 publications

(12 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Wen et al [46] combine the folding module with self-attention in a hierarchical decoder that makes use of features extracted from the multi-level encoder. New encoders are proposed by [44] and [58] to learn better local features from neighbor points for point cloud completion and other tasks such as segmentation. PMP-Net [47] generates complete point clouds by moving input points to appropriate positions iteratively with minimum moving distance.…”

Section: Related Workmentioning

confidence: 99%

“…Our specially designed CS-Blocks also make the network refine the segmentation and completion results block by block. For the third evaluation way, we use PSV [58], PointNet++ [33] and Point Transformer [62] as segmentation networks and SnowflakeNet [49] and PMP-Net [47] as completion networks for fusion, as shown in Table 3. Because of the direct feature sharing between segmentation and completion in these end-to-end networks, the shared features need to offer the information for both segmentation and completion, which is hard for the network to learn, leading to the bad performance of the end-to-end framework.…”

Section: Comparisonmentioning

confidence: 99%

See 1 more Smart Citation

Completing Partial Point Clouds with Outliers by Collaborative Completion and Segmentation

Ma¹,

Yang²,

Guo³

et al. 2022

Preprint

View full text Add to dashboard Cite

Most existing point cloud completion methods are only applicable to partial point clouds without any noises and outliers, which does not always hold in practice. We propose in this paper an end-toend network, named CS-Net, to complete the point clouds contaminated by noises or containing outliers. In our CS-Net, the completion and segmentation modules work collaboratively to promote each other, benefited from our specifically designed cascaded structure. With the help of segmentation, more clean point cloud is fed into the completion module. We design a novel completion decoder which harnesses the labels obtained by segmentation together with FPS to purify the point cloud and leverages KNN-grouping for better generation. The completion and segmentation modules work alternately share the useful information from each other to gradually improve the quality of prediction. To train our network, we build a dataset to simulate the real case where incomplete point clouds contain outliers. Our comprehensive experiments and comparisons against state-of-the-art completion methods demonstrate our superiority. We also compare with the scheme of segmentation followed by completion and their end-to-end fusion, which also proves our efficacy.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Comparisonmentioning

confidence: 99%

Completing Partial Point Clouds with Outliers by Collaborative Completion and Segmentation

Ma¹,

Yang²,

Guo³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…For more details, we pass the 𝑃 𝑠𝑦𝑛𝑡ℎ𝑒𝑡𝑖𝑐 through a series of bottomup and top-down structural styles of MLPs. Each point in 𝑃 𝑠𝑦𝑛𝑡ℎ𝑒𝑡𝑖𝑐 is firstly encoded into 𝐹 1 with multiple dimensions [128,64]. Then we decode it to 𝐹 2 passing through MLPs with multiple dimensions [64,128,64].…”

Section: Refinement Unitmentioning

confidence: 99%

“…Each point in 𝑃 𝑠𝑦𝑛𝑡ℎ𝑒𝑡𝑖𝑐 is firstly encoded into 𝐹 1 with multiple dimensions [128,64]. Then we decode it to 𝐹 2 passing through MLPs with multiple dimensions [64,128,64]. To preserve 𝐹 1 in the following layers, we combine 𝐹 1 with 𝐹 2 , as shown in Fig.…”

Section: Refinement Unitmentioning

confidence: 99%

See 1 more Smart Citation

ASFM-Net: Asymmetrical Siamese Feature Matching Network for Point Completion

Xia

Wei³

et al. 2021

Preprint

View full text Add to dashboard Cite

We tackle the problem of object completion from point clouds and propose a novel point cloud completion network employing an asymmetrical Siamese feature matching strategy, termed as ASFM-Net. Specifically, the asymmetrical Siamese auto-encoder neural network is adopted to map the partial and complete input point cloud into a shared latent space, which can capture detailed shape prior. Then we design an iterative refinement unit to generate complete shapes with fine-grained details by integrating prior information. Experiments are conducted on PCN dataset and Completion3D benchmark, demonstrating the state-of-the-art performance of the proposed ASFM-Net. Our method achieves the 1st place in the leaderboard of Completion3D 1 and outperforms existing methods with a large margin, about 12%. Codes will be open-sourced.

show abstract

Multi-space and detail-supplemented attention network for point cloud completion

Xiang

Yang

et al. 2022

Appl Intell

View full text Add to dashboard Cite

Point Set Voting for Partial Point Cloud Analysis

Cited by 31 publications

References 28 publications

Completing Partial Point Clouds with Outliers by Collaborative Completion and Segmentation

Completing Partial Point Clouds with Outliers by Collaborative Completion and Segmentation

ASFM-Net: Asymmetrical Siamese Feature Matching Network for Point Completion

Multi-space and detail-supplemented attention network for point cloud completion

Contact Info

Product

Resources

About