Hypergraph Convolutional Network Based Weakly Supervised Point Cloud Semantic Segmentation with Scene-Level Annotations

Lu, Zhuheng; Zhang, Peng; Dai, Yuewei; Li, Weiqing; Su, Zhiyong

doi:10.2139/ssrn.4348441

Cited by 2 publications

(13 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Yang et al [21] combined the original point cloud with a downsampled point cloud that underwent random point sampling to use inter-cloud semantics as supervision. Liu et al [24] introduced a Region-wise Masking (Region-Mask) strategy with augmented data to get masked point cloud, which contains meaningful context.…”

Section: Methodsmentioning

confidence: 99%

“…It uses manifold learning to optimise the selection of initial weak annotations to retain more significant semantic data by projecting the extracted features to a more suitable feature space for combination. Influenced by Superpoint Graphs [49], Cheng et al [28] and Lu et al [41] decided to use the original point cloud to generate the superpoints and construct superpoint graphs to mine the long-range dependencies and balance the point numbers among different classes.…”

Section: Methodsmentioning

confidence: 99%

“…The first stage gets a small number of high-quality pseudo labels by temporal matching and uses them to train the initial model, and the second stage utilises the first model to extract features and combines temporal matching and spatial graph propagation to update the pseudo labels. Lu et al [41] designed a novel Weighted Hypergraph Convolutional Network (WHCN), which consists of a hyperedge attention module and a spectral hypergraph convolution module. Weighted Hypergraph Convolutional Network aggregated the features and captured highorder semantic relations to generate accurate point-level pseudo labels.…”

Section: Generate Pseudo Labelsmentioning

confidence: 99%

“…Wu et al [37] proposed a Dual Adaptive Transformations model based on KPConv, which improves segmentation performance using generated perturbed and deformed samples to enhance the local and structural consistency constraints of 3D point clouds, thus improving the segmentation performance. Liu et al [24] utilised consistency constraints between different augmented data and masked data to learn the complex scene context. Wu et al [42] designed a separation and consistency module to apply consistency constraints with reliable and ambiguous pseudo labels, a mixaugmentation module to further enforce the consistency constraints for the reliable points.…”

Section: Semantic Refinementmentioning

confidence: 99%

“…Yao et al [26] created limitations by balancing the authenticity of pseudo labels and original labels. Using self-supervised learning and label propagation, Zhang et al [27] restricted and optimised the network; Cheng et al [28] added pseudo label and attentional feature constraints; Wang and Yao [30,33], Liu et al [31,43], Shi et al [35] and Lu et al [41] all incorporated consistency and pseudo label constraints. As a novel hybrid constraint, Kong et al [36] merged pseudo labelling with weakly supervised information.…”

Section: Semantic Refinementmentioning

confidence: 99%

See 4 more Smart Citations

A survey on weakly supervised 3D point cloud semantic segmentation

Wang,

Liu,

Tan

et al. 2023

IET Computer Vision

View full text Add to dashboard Cite

With the popularity and advancement of 3D point cloud data acquisition technologies and sensors, research into 3D point clouds has made considerable strides based on deep learning. The semantic segmentation of point clouds, a crucial step in comprehending 3D scenes, has drawn much attention. The accuracy and effectiveness of fully supervised semantic segmentation tasks have greatly improved with the increase in the number of accessible datasets. However, these achievements rely on time‐consuming and expensive full labelling. In solve of these existential issues, research on weakly supervised learning has recently exploded. These methods train neural networks to tackle 3D semantic segmentation tasks with fewer point labels. In addition to providing a thorough overview of the history and current state of the art in weakly supervised semantic segmentation of 3D point clouds, a detailed description of the most widely used data acquisition sensors, a list of publicly accessible benchmark datasets, and a look ahead to potential future development directions is provided.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Generate Pseudo Labelsmentioning

confidence: 99%

Section: Semantic Refinementmentioning

confidence: 99%

Section: Semantic Refinementmentioning

confidence: 99%

See 3 more Smart Citations

A survey on weakly supervised 3D point cloud semantic segmentation

Wang,

Liu,

Tan

et al. 2023

IET Computer Vision

View full text Add to dashboard Cite

show abstract

RM3D: Robust Data-Efficient 3D Scene Parsing via Traditional and Learnt 3D Descriptors-Based Semantic Region Merging

Liu

2022

Int J Comput Vis

View full text Add to dashboard Cite

Existing state-of-the-art 3D point cloud understanding methods merely perform well in a fully supervised manner. To the best of our knowledge, there exists no unified framework that simultaneously solves the downstream high-level understanding tasks including both segmentation and detection, especially when labels are extremely limited. This work presents a general and simple framework to tackle point cloud understanding when labels are limited. The first contribution is that we have done extensive methodology comparisons of traditional and learned 3D descriptors for the task of weakly supervised 3D scene understanding, and validated that our adapted traditional PFH-based 3D descriptors show excellent generalization ability across different domains. The second contribution is that we proposed a learning-based region merging strategy based on the affinity provided by both the traditional/learned 3D descriptors and learned semantics. The merging process takes both low-level geometric and high-level semantic feature correlations into consideration. Experimental results demonstrate that our framework has the best performance among the three most important weakly supervised point clouds understanding tasks including semantic segmentation, instance segmentation, and object detection even when very limited number of points are labeled. Our method, termed Region Merging 3D (RM3D), has superior performance on Scan-Net data-efficient learning online benchmarks and other four large-scale 3D understanding benchmarks under various experimental settings, outperforming current arts by a margin for various 3D understanding tasks without complicated learning strategies such as active learning.

show abstract

Hypergraph Convolutional Network Based Weakly Supervised Point Cloud Semantic Segmentation with Scene-Level Annotations

Cited by 2 publications

References 0 publications

A survey on weakly supervised 3D point cloud semantic segmentation

A survey on weakly supervised 3D point cloud semantic segmentation

RM3D: Robust Data-Efficient 3D Scene Parsing via Traditional and Learnt 3D Descriptors-Based Semantic Region Merging

Contact Info

Product

Resources

About