Quality-Aware Compression of Point Clouds with Google Draco

Hoog, Jens de; Ahmed, Ahmed N.; Anwar, Ali; Latré, Steven; Hellinckx, Peter

doi:10.1007/978-3-030-89899-1_23

Cited by 10 publications

(5 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The tree data structure is a memory-efficient representation for sparse point clouds. Draco [4] and MPEG G-PCC [5] are wellknown tree-based methods with KD-tree and octree data structures, respectively, but neither uses deep-learned approaches. Recently, several deep-learned octree compression techniques have emerged.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

University Technology Commercialization in Taiwan: National Taiwan University (NTU) and National Taiwan University of Science and Technology (NTUST)

Liu¹,

Chen²,

Chiou³

et al. 2011

Academic Entrepreneurship in Asia

View full text Add to dashboard Cite

The large amount of data collected by LiDAR sensors brings the issue of LiDAR point cloud compression (PCC). Previous works on LiDAR PCC have used range image representations and followed the predictive coding paradigm to create a basic prototype of a coding framework. However, their prediction methods give an inaccurate result due to the negligence of invalid pixels in range images and the omission of future frames in the time step. Moreover, their handcrafted design of residual coding methods could not fully exploit spatial redundancy. To remedy this, we propose a coding framework BIRD-PCC. Our prediction module is aware of the coordinates of invalid pixels in range images and takes a bidirectional scheme. Also, we introduce a deep-learned residual coding module that can further exploit spatial redundancy within a residual frame. Experiments conducted on SemanticKITTI and KITTI-360 datasets show that BIRD-PCC outperforms other methods in most bitrate conditions and generalizes well to unseen environments.

show abstract

Section: Related Workmentioning

confidence: 99%

“…Two tree-based baselines and one range imagebased baseline are selected in experiments. Draco [4] is an open-source compression algorithm based on kd-tree released by Google. G-PCC [5] is a compression standard based on octree proposed by MPEG.…”

Section: Settingsmentioning

confidence: 99%

University Technology Commercialization in Taiwan: National Taiwan University (NTU) and National Taiwan University of Science and Technology (NTUST)

Liu¹,

Chen²,

Chiou³

et al. 2011

Academic Entrepreneurship in Asia

View full text Add to dashboard Cite

show abstract

“…According to some legal scholars force majeure events or accidents must be external (Ripert 1922; Berlingieri 2014) a sudden failure of the autonomous system could not be considered a force majeure event, similarly, a cyber-attack provoking the to the system cannot be deemed as an event unavoidable or unexpected with due diligence in the context of autonomous shipping. However, different opinion is upheld by other scholars (Rodière 1972;Waldstein et al 2007;De Decker 2015). Accordingly, the unexpected breaking of machinery onboard-e.g.…”

Section: Geneva Collision Convention 1960mentioning

confidence: 96%

Autonomous inland navigation: a literature review and extracontractual liability issues

Domenighini

2024

J. shipp. trd.

View full text Add to dashboard Cite

Pilot tests for (semi-)autonomous transport via inland waterways are already taking place in Belgium and other European countries. However, the full commercial implementation of autonomous inland shipping might be hampered by liability issues. The allocation of liability, especially extracontractual liability, is an important concern for shipowners who want to invest in autonomous barges, and for other ecosystem actors. For this reason, a balanced risk distribution framework could boost the investment decision. A legal analysis of the current laws and regulations is necessary to evaluate whether they can be applied to new autonomous systems. The research approach consists of two steps. First, an in-depth literature review is carried out to determine whether extracontractual liability research has already been conducted and to highlight the gaps in autonomous inland waterways transport studies. Once it is proved that the vast majority of the research focuses on technology, it can be affirmed that there is a need to inquire about extracontractual liability. In the second part, thus, the two international Conventions on liability limitation and collision in inland navigation are examined.

show abstract

“…Huang et al proposed a 3D point-cloud geometric compression method based on deep learning, which was an auto-encoder that also supported the parallel compression of multiple models via the GPU, considerably increasing processing efficiency [ 23 ]. When compared to PCL compression [ 24 ] and Draco compression [ 25 ], this method retains the original shape with little loss. To learn how to represent disordered point clouds, Panos et al constructed an auto-encoder using convolutional layers and fully linked layers [ 26 ].…”

Section: Related Workmentioning

confidence: 99%

An Optimized Convolutional Neural Network for the 3D Point-Cloud Compression

Luo

Xiong

et al. 2023

Sensors

View full text Add to dashboard Cite

Due to the tremendous volume taken by the 3D point-cloud models, knowing how to achieve the balance between a high compression ratio, a low distortion rate, and computing cost in point-cloud compression is a significant issue in the field of virtual reality (VR). Convolutional neural networks have been used in numerous point-cloud compression research approaches during the past few years in an effort to progress the research state. In this work, we have evaluated the effects of different network parameters, including neural network depth, stride, and activation function on point-cloud compression, resulting in an optimized convolutional neural network for compression. We first have analyzed earlier research on point-cloud compression based on convolutional neural networks before designing our own convolutional neural network. Then, we have modified our model parameters using the experimental data to further enhance the effect of point-cloud compression. Based on the experimental results, we have found that the neural network with the 4 layers and 2 strides parameter configuration using the Sigmoid activation function outperforms the default configuration by 208% in terms of the compression-distortion rate. The experimental results show that our findings are effective and universal and make a great contribution to the research of point-cloud compression using convolutional neural networks.

show abstract

Quality-Aware Compression of Point Clouds with Google Draco

Cited by 10 publications

References 16 publications

University Technology Commercialization in Taiwan: National Taiwan University (NTU) and National Taiwan University of Science and Technology (NTUST)

University Technology Commercialization in Taiwan: National Taiwan University (NTU) and National Taiwan University of Science and Technology (NTUST)

Autonomous inland navigation: a literature review and extracontractual liability issues

An Optimized Convolutional Neural Network for the 3D Point-Cloud Compression

Contact Info

Product

Resources

About