RFNet: Recurrent Forward Network for Dense Point Cloud Completion

Huang, Tianxin; Zou, Hao; Cui, Jinhao; Yang, Xin; Wang, Mengmeng; Zhao, Xiangrui; Zhang, Jiangning; Xu, Yifan; Liu, Yong

doi:10.1109/iccv48922.2021.01228

Cited by 29 publications

(8 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This method achieves a high-fidelity dense point cloud completion through decoding a complete but sparse shape, iterative refinement, preserving trustworthy information by symmetrization, and patch-wise up-sampling. Recently, a Recurrent Forward Network (RFNet) consisting of three modules was devised by Huang et al [42], where there are Recurrent Feature Extraction (RFE) module, Forward Dense Completion (FDC) module, and Raw Shape Protection (RSP) module. The RFE extracts multiple global features from the incomplete point clouds for different recurrent levels, while the FDC produces the output in a coarse-to-fine pipeline.…”

Section: A Point-based Methodsmentioning

confidence: 99%

Comprehensive Review of Deep Learning-Based 3D Point Cloud Completion Processing and Analysis

Fei¹,

Yang²,

Chen³

et al. 2022

Preprint

View full text Add to dashboard Cite

Point cloud completion is a generation and estimation issue derived from the partial point clouds, which plays a vital role in the applications in 3D computer vision. The progress of deep learning (DL) has impressively improved the capability and robustness of point cloud completion. However, the quality of completed point clouds is still needed to be further enhanced to meet the practical utilization. Therefore, this work aims to conduct a comprehensive survey on various methods, including point-based, convolution-based, graph-based, and generative model-based approaches, etc. And this survey summarizes the comparisons among these methods to provoke further research insights. Besides, this review sums up the commonly used datasets and illustrates the applications of point cloud completion. Eventually, we also discussed possible research trends in this promptly expanding field.

show abstract

Section: A Point-based Methodsmentioning

confidence: 99%

Comprehensive Review of Deep Learning-Based 3D Point Cloud Completion Processing and Analysis

Fei¹,

Yang²,

Chen³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Unlike the methods which are dependent on a vectorized global feature to solve the permutation invariant problem, RFNet (Huang et al 2021) and PointTr (Yu et al 2021) produce several global features in their encoder. On one hand, RFNet (Huang et al 2021) uses their features to complete the object in an recurrent way by concatenating the incomplete input and the predicted points level by level. On the other, PointTr (Yu et al 2021) relies on transformers to produce a set of queries directly from the observed points with the help of positional coding.…”

Section: Point Cloudmentioning

confidence: 99%

SoftPool++: An Encoder–Decoder Network for Point Cloud Completion

et al. 2022

View full text Add to dashboard Cite

We propose a novel convolutional operator for the task of point cloud completion. One striking characteristic of our approach is that, conversely to related work it does not require any max-pooling or voxelization operation. Instead, the proposed operator used to learn the point cloud embedding in the encoder extracts permutation-invariant features from the point cloud via a soft-pooling of feature activations, which are able to preserve fine-grained geometric details. These features are then passed on to a decoder architecture. Due to the compression in the encoder, a typical limitation of this type of architectures is that they tend to lose parts of the input shape structure. We propose to overcome this limitation by using skip connections specifically devised for point clouds, where links between corresponding layers in the encoder and the decoder are established. As part of these connections, we introduce a transformation matrix that projects the features from the encoder to the decoder and vice-versa. The quantitative and qualitative results on the task of object completion from partial scans on the ShapeNet dataset show that incorporating our approach achieves state-of-the-art performance in shape completion both at low and high resolutions.

show abstract

“…(1) Traditional point cloud completion methods [15], [16], [17], [18] usually assume a smooth surface of 3D shape, or utilize large-scaled complete shape dataset to infer the missing regions for incomplete shapes. (2) Deep learning based methods [9], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39], [40], [41], [42], [43], [44], however, learn to predict a complete shape based on the learned prior from the training data. Our method falls into the second category and focuses on the decoding process of point cloud completion.…”

Section: Related Workmentioning

confidence: 99%

Snowflake Point Deconvolution for Point Cloud Completion and Generation with Skip-Transformer

Peng¹,

Xu²,

Liu³

et al. 2022

Preprint

View full text Add to dashboard Cite

Most existing point cloud completion methods suffered from discrete nature of point clouds and unstructured prediction of points in local regions, which makes it hard to reveal fine local geometric details. To resolve this issue, we propose SnowflakeNet with Snowflake Point Deconvolution (SPD) to generate the complete point clouds. SPD models the generation of complete point clouds as the snowflake-like growth of points, where the child points are progressively generated by splitting their parent points after each SPD. Our insight of revealing detailed geometry is to introduce skip-transformer in SPD to learn point splitting patterns which can fit local regions the best. Skip-transformer leverages attention mechanism to summarize the splitting patterns used in previous SPD layer to produce the splitting in current SPD layer. The locally compact and structured point clouds generated by SPD precisely reveal the structure characteristic of 3D shape in local patches, which enables us to predict highly detailed geometries. Moreover, since SPD is a general operation, which is not limited to completion, we further explore the applications of SPD on other generative tasks, including point cloud auto-encoding, generation, single image reconstruction and upsampling. Our experimental results outperform the state-of-the-art methods under widely used benchmarks.

show abstract

RFNet: Recurrent Forward Network for Dense Point Cloud Completion

Cited by 29 publications

References 26 publications

Comprehensive Review of Deep Learning-Based 3D Point Cloud Completion Processing and Analysis

Comprehensive Review of Deep Learning-Based 3D Point Cloud Completion Processing and Analysis

SoftPool++: An Encoder–Decoder Network for Point Cloud Completion

Snowflake Point Deconvolution for Point Cloud Completion and Generation with Skip-Transformer

Contact Info

Product

Resources

About