Deep Implicit Volume Compression

  title={Deep Implicit Volume Compression},
  author={Danhang Tang and Saurabh Singh and Philip A. Chou and Christian Haene and Mingsong Dou and S. Fanello and Jonathan Taylor and Philip L. Davidson and Onur G. Guleryuz and Yinda Zhang and Shahram Izadi and Andrea Tagliasacchi and Sofien Bouaziz and Cem Keskin},
  journal={2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  • Danhang Tang, Saurabh Singh, +11 authors Cem Keskin
  • Published 18 May 2020
  • Engineering, Computer Science
  • 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
We describe a novel approach for compressing truncated signed distance fields (TSDF) stored in 3D voxel grids, and their corresponding textures. To compress the TSDF, our method relies on a block-based neural network architecture trained end-to-end, achieving state-of-the-art rate-distortion trade-off. To prevent topological errors, we losslessly com- press the signs of the TSDF, which also upper bounds the reconstruction error by the voxel size. To compress the corresponding texture, we… Expand
3D Shapes Local Geometry Codes Learning with SDF
  • Shun Yao, Fei Yang, Yongmei Cheng, M. Mozerov
  • Computer Science
  • 2021 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW)
  • 2021
This paper proposes Local Geometry Code Learning (LGCL), a model that improves the original DeepSDF results by learning from a local shape geometry of the full 3D shape by adding an extra graph neural network to split the single transmittable latent code into a set of local latent codes distributed on the3D shape. Expand
DeepCompress: Efficient Point Cloud Geometry Compression
Through an ablation study, it is shown that incorporating the learned activation function from Computational Efficient Neural Image Compression and designing more parameter-efficient convolutional blocks yields dramatic gains in efficiency and performance. Expand
Neural Geometric Level of Detail: Real-time Rendering with Implicit 3D Shapes
An efficient neural representation is introduced that enables real-time rendering of high-fidelity neural SDFs, while achieving state-of-the-art geometry reconstruction quality, and is 2–3 orders of magnitude more efficient in terms of rendering speed. Expand
Multiscale Point Cloud Geometry Compression
This paper proposes a multiscale end-to-end learning framework that hierarchically reconstructs the 3D Point Cloud Geometry (PCG) via progressive re-sampling on top of a sparse convolution based autoencoder for point cloud compression and reconstruction. Expand
Multiresolution Deep Implicit Functions for 3D Shape Representation
Multiresolution Deep Implicit Functions (MDIF) is introduced, a hierarchical representation that can recover fine geometry detail, while being able to perform global operations such as shape completion and support both encoder-decoder inference and decoder-only latent optimization. Expand
A deep perceptual metric for 3D point clouds
This paper proposes a perceptual loss function for 3D point clouds which outperforms existing loss functions on the ICIP2020 subjective dataset and proposes a novel truncated distance field voxel grid representation which leads to sparser latent spaces and loss functions that are more correlated with perceived visual quality compared to a binary representation. Expand
Better Patch Stitching for Parametric Surface Reconstruction
It is shown on standard benchmarks that the use of normal consistency requirement outperforms the baselines quantitatively while enforcing better stitching leads to much better visual quality of the reconstructed objects as compared to the state-of-the-art. Expand
LVAC: Learned Volumetric Attribute Compression for Point Clouds using Coordinate Based Networks
This is the first work to compress volumetric functions represented by local coordinate-based neural networks, and expects it to be applicable beyond point clouds, for example to compression of high-resolution neural radiance fields. Expand
Adaptive Deep Learning-Based Point Cloud Geometry Coding
A novel deep learning-based solution for point cloud geometry coding which is able to efficiently adapt to the content's characteristics and maximizing the compression performance is presented. Expand
Trajectory Prediction with Compressed 3D Environment Representation using Tensor Train Decomposition
Trajectory optimization for motion planning and optimal control is a popular approach in robotics. Algorithms typically require good initialization in order to find the optimal trajectories. ToExpand


Real-time compression and streaming of 4D performances
We introduce a realtime compression architecture for 4D performance capture that is two orders of magnitude faster than current state-of-the-art techniques, yet achieves comparable visual quality andExpand
Learning Convolutional Transforms for Lossy Point Cloud Geometry Compression
This paper presents a novel data-driven geometry compression method for static point clouds based on learned convolutional transforms and uniform quantization that outperforms the MPEG reference solution in terms of rate-distortion on the Microsoft Voxelized Upper Bodies dataset. Expand
Variational image compression with a scale hyperprior
It is demonstrated that this model leads to state-of-the-art image compression when measuring visual quality using the popular MS-SSIM index, and yields rate-distortion performance surpassing published ANN-based methods when evaluated using a more traditional metric based on squared error (PSNR). Expand
Deep Marching Cubes: Learning Explicit Surface Representations
This paper demonstrates that the marching cubes algorithm is not differentiable and proposes an alternative differentiable formulation which is inserted as a final layer into a 3D convolutional neural network, and proposes a set of loss functions which allow for training the model with sparse point supervision. Expand
End-to-end Optimized Image Compression
Across an independent set of test images, it is found that the optimized method generally exhibits better rate-distortion performance than the standard JPEG and JPEG 2000 compression methods, and a dramatic improvement in visual quality is observed, supported by objective quality estimates using MS-SSIM. Expand
Octree-based Point-Cloud Compression
A progressive compression method for point sampled models that is specifically apt at dealing with densely sampled surface geometry and it is demonstrated that additional point attributes, such as color, can be well integrated and efficiently encoded in this framework. Expand
DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation
This work introduces DeepSDF, a learned continuous Signed Distance Function (SDF) representation of a class of shapes that enables high quality shape representation, interpolation and completion from partial and noisy 3D input data. Expand
Deep AutoEncoder-based Lossy Geometry Compression for Point Clouds
This paper proposes the first autoencoder-based geometry compression codec that directly takes point clouds as input rather than voxel grids or collections of images, and shows that the proposed method outperforms the test model for categories 1 and 3 published by MPEG-3DG group on the 125th meeting. Expand
Compression of 3D Point Clouds Using a Region-Adaptive Hierarchical Transform
  • R. Queiroz, P. Chou
  • Computer Science, Medicine
  • IEEE Transactions on Image Processing
  • 2016
This paper devise a method to compress the colors in point clouds, which is based on a hierarchical transform and arithmetic coding, and believes this paper represents the state of the art in intra-frame compression of point clouds for real-time 3D video. Expand
TFAN: A low complexity 3D mesh compression algorithm
This paper proposes a novel approach for mono‐resolution 3D mesh compression, called TFAN (Triangle Fan‐based compression), which treats in a unified manner meshes of arbitrary topologies while offering a linear computational complexity for both encoding and decoding algorithms. Expand