Learning residual coding for point clouds

Lazzarotto, Davi; Ebrahimi, Touradj

doi:10.1117/12.2597814

Cited by 6 publications

(4 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Drawing inspiration from the remarkable achievements of learning-based methods in image and video compression, similar architectures have been adopted for PCGC. For static PCGC, early works employed dense 3D convolutions in autoencoder architectures for lossy PCGC [2,3,4] and block prediction [5]. Alternatively, voxel occupancy values were directly estimated [6] for lossless coding approaches.…”

Section: Related Workmentioning

confidence: 99%

Temporal Conditional Coding for Dynamic Point Cloud Geometry Compression

Huang,

Lazzarotto,

Ebrahimi

2024

ICASSP 2024 - 2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

Point clouds allow for the representation of 3D multimedia content as a set of disconnected points in space. Their inherent irregular geometric nature poses a challenge to efficient compression, a critical operation for both storage and transmission. This paper proposes a VAE-inspired codec tailored for dynamic point cloud geometry compression, taking advantage of a temporal autoregressive hyperprior to enhance compression performance. Specifically, features derived from adjacent point cloud frames help build a hyperprior for conditional entropy coding. Sparse convolutions are leveraged to reach higher computational efficiency when compared to 3D dense convolutions. Remarkably, the proposed approach achieves an average 60.2% BD-rate gain against the contemporary V-PCC compression standard from MPEG.

show abstract

Section: Related Workmentioning

confidence: 99%

Temporal Conditional Coding for Dynamic Point Cloud Geometry Compression

Huang,

Lazzarotto,

Ebrahimi

2024

ICASSP 2024 - 2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

show abstract

“…An autoregressive entropy coding model was also explored with comparable architecture [15], producing even better results that outperformed V-PCC for the evaluated test set. Other techniques, such as block prediction [16] and residual coding [17] explored further extension of similar techniques. Recently, the JPEG standardisation committee launched a call for proposals for learning-based point cloud coding, and a compression method based on dense convolutions was selected as the starting point known by the term verification model [8].…”

Section: Related Workmentioning

confidence: 99%

Evaluating the effect of sparse convolutions on point cloud compression

Lazzarotto,

Ebrahimi

2023

2023 11th European Workshop on Visual Information Processing (EUVIP)

Self Cite

View full text Add to dashboard Cite

The use of point clouds as an imaging modality has been rapidly growing, motivating research on compression methods to enable efficient transmission and storage for many applications. While compression standards relying on conventional techniques such as planar projection and octree-based representation have been standardized by MPEG, recent research has demonstrated the potential of neural networks in achieving better rate-distortion performance for point cloud geometry coding. Early attempts in learning-based point cloud coding mostly relied on autoencoder architectures using dense convolutional layers, but the majority of recent research has shifted towards the use of sparse convolutions, which are applied only to occupied positions rather than the entire space. Since points are usually distributed on underlying surfaces rather than volumes, such operations allow to reduce the computational complexity required to compress and decompress point clouds. Moreover, recent solutions also achieve better compression efficiency, allocating fewer bits at similar levels of geometric distortion. However, it is not clear to which extent this gain in performance is due to the use of sparse convolutions, if any at all, since the architecture of the model is often modified. In this paper, we conduct an evaluation of the effect of replacing dense convolutions with sparse convolutions on the rate-distortion performance of the JPEG Pleno Point Cloud Verification Model. Results show that the use of sparse convolutions allows for an average BD-rate reduction of approximately 9% for both D1 and D2 PSNR metrics based on similar training procedures, with an even bigger reduction in point clouds featuring reduced point density.

show abstract

“…Learning-based versions of methods popular in common video codecs were also studied, such as intra prediction 22 and residual coding. 23 An important development was introduced by Wang et al, 24 which used sparse convolutions in order to take better advantage of the fact that the majority of voxels in point clouds are empty, operating only on positions that are effectively occupied by points. Sparse convolutions were further explored by the same authors 25 using cross-scale and cross-stage prediction to achieve lossless compression as well as superior lossy rate-distortion performance.…”

Section: Geometry Compressionmentioning

confidence: 99%

Integrated learning-based point cloud compression for geometry and color with graph Fourier transforms

Lazzarotto

Ebrahimi

2022

Applications of Digital Image Processing XLV

Self Cite

View full text Add to dashboard Cite

Point cloud representation is a popular modality to code immersive 3D contents. Several solutions and standards have been recently proposed in order to efficiently compress the large volume of data that point clouds require, in order to make them feasible for real-life applications. Recent studies adopting learning-based methods for point cloud compression have demonstrated high compression efficiency specially when compared to the conventional compression standards. However, they are mostly evaluated either on geometry or color separately, and few learning-based joint codecs have been proposed. In this paper, we propose an integrated learned coding architecture by joining a previously proposed geometry coding module based on three-dimensional convolutional layers with a color compression method relying on graph Fourier transform (GFT) using a learning-based mean and scale hyperprior to compress the obtained coefficients. Evaluation on a test set with dense point clouds shows that the proposed method outperforms G-PCC and achieves competitive performance with V-PCC when evaluated with state-of-the-art objective quality metrics.

show abstract

Learning residual coding for point clouds

Cited by 6 publications

References 38 publications

Temporal Conditional Coding for Dynamic Point Cloud Geometry Compression

Temporal Conditional Coding for Dynamic Point Cloud Geometry Compression

Evaluating the effect of sparse convolutions on point cloud compression

Integrated learning-based point cloud compression for geometry and color with graph Fourier transforms

Contact Info

Product

Resources

About