Problem Of Compression Research Articles

Optimizing H.266/VVC Intra Coding with a Genetic Algorithm: Balancing Speed and Quality

The growing need for high-definition video material requires improvements in video encoding systems that maximize encoding performance while simultaneously improving compression efficiency. This paper presents a novel genetic algorithm-based intra-coding optimization method for the H.266Versatile Video Coding (VVC) standard. One of the biggest problems in video compression is finding the ideal balance between encoding speed and video quality, which is what our approach aims to solve. Our suggested method makes use of the strong search capabilities of the evolutionary algorithm to choose the best Multi-Type Tree (MTT) partitions and coding tools from the wide range of possibilities present in H.266VVC. The wellness assessment work that guides this choice method combines criteria for perceptual appraisal of video quality and measures for coding productivity appraisal.

Distance-Preserving Graph Compression Techniques

We study the problem of distance-preserving graph compression for weighted paths and trees. The problem entails a weighted graph $G = (V, E)$ with non-negative weights and a subset of edges $E^{\prime} \subset E$, which needs to be removed from G (with their endpoints merged as a supernode). The goal is to redistribute the weights of the deleted edges in a way that minimizes the error. The error is defined as the sum of the absolute differences of the shortest path lengths between different pairs of nodes before and after contracting $E^{\prime}$. Based on this error function, we propose optimal approaches for merging any subset of edges in a path and a single edge in a tree. Previous works on graph compression techniques aimed at preserving different graph properties (such as the chromatic number) or solely focused on identifying the optimal set of edges to contract. However, our focus in this paper is on achieving optimal edge contraction (when the contracted edges are provided as input), specifically for weighted trees and paths.

QARV: Quantization-Aware ResNet VAE for Lossy Image Compression.

This paper addresses the problem of lossy image compression, a fundamental problem in image processing and information theory that is involved in many real-world applications. We start by reviewing the framework of variational autoencoders (VAEs), a powerful class of generative probabilistic models that has a deep connection to lossy compression. Based on VAEs, we develop a new scheme for lossy image compression, which we name quantization-aware ResNet VAE (QARV). Our method incorporates a hierarchical VAE architecture integrated with test-time quantization and quantization-aware training, without which efficient entropy coding would not be possible. In addition, we design the neural network architecture of QARV specifically for fast decoding and propose an adaptive normalization operation for variable-rate compression. Extensive experiments are conducted, and results show that QARV achieves variable-rate compression, high-speed decoding, and better rate-distortion performance than existing baseline methods.

Enhancing the Compaction Characteristics of Peat Soil through Ground Granulated Blast Furnace Slag (GGBS) Stabilisation

Peat soil is renowned for its compressibility problem that can give rise to issues such as ground settlement, particularly within the spectrum of construction endeavours. Conversely, the implementation of chemical stabilisation, involving the incorporation of suitable additives, is contemplated prior to commencing construction activities. Ground Granulated Blast Furnace Slag (GGBS), which is a byproduct of the iron and steel industry, stands out as a prominent additive that can augment the engineering attributes of peat soil. This investigation focuses on the compaction characteristics of peat soil treated with varying proportions of GGBS, as evaluated through the Modified Proctor test. It involved examining the impact of diverse GGBS additions on the liquid limit of peat soil. The findings showed that the introduction of GGBS yielded a substantial reduction in the liquid limit of peat soil, signifying an elevated susceptibility to moisture and an increase in plasticity. Moreover, the incorporation of GGBS into peat soil prompted an increase in its Maximum Dry Density (MDD) and a reduction in its Optimum Moisture Content (OMC). Notably, the peak MDD, registering at 1000 kg/m³, was achieved at a GGBS content of 50%, marking the highest value observed in this study. The OMC values further indicated the lowest moisture content, standing at 29%, with the addition of 25% GGBS to the peat soil. These outcomes convincingly advocate that the utilisation of GGBS in treating peat soil engenders positive enhancement in its compaction characteristics, specifically pertaining to MDD and OMC. Such enhancement in engineering properties holds promising implications for potential stability.

Extensional flow of a compressible viscous fluid

We derive reduced models for extrusion problems where it is necessary to account for fluid compressibility. We consider the two-dimensional extensional flow of a compressible viscous fluid and discuss two specific applications: weakly compressible fluids and bubbly liquid–gas mixtures that behave as a single compressible fluid. The mathematical model we present consists of equations for conservation of mass, conservation of momentum and a closure relationship between the pressure and density. The most substantial differences between compressible extrusion problems is in the closure relationship. By integrating the conservation equations across the fluid cross-section and exploiting a slender aspect ratio, we derive reduced equations for conservation of mass and conservation of momentum in the direction of flow. The reduced system of equations relating cross-sectionally averaged quantities is closed by a relationship between the averaged pressure and density, which will differ substantially depending on the application. We demonstrate the utility of a reduced model for both the weakly compressible fluid and bubbly mixture applications; namely, in providing valuable quantitative insights without needing to solve a complicated free-boundary problem.

A parameter-free staggered-grid Lagrangian scheme for two-dimensional compressible flow problems

The present study aims to develop a parameter-free staggered-grid Lagrangian scheme that avoids the empirical parameters needed for artificial viscosity and anti-hourglass force. The artificial viscosity is equal to the pressure jump, which is constructed by the relationship between pressure and velocity. The anti-hourglass force is established using pressure compensation, which is the difference in pressure between sub-cells belonging to the same primary cell. The scheme maintains the conservation of total mass, momentum, and energy. Numerical experiments show that the scheme is highly robust for extreme flow problems with different fine meshes such as Sedov, Noh, Saltzmann triple point and Rayleigh–Taylor instability. The robust parameter-free scheme is well suited for multi-physics problems and engineering applications.

Fast Method Of Image File Compression In Web-Based Application Form Using Huffman Algorithm

Many web-based applications/forms serve as online data entry devices equipped with facilities for uploading image data. When uploading image data, the image file is often incompatible or cannot be uploaded because the size is too large. Users have to resize those files separately. For example, with several compression applications used such as photoshop, gimp, photoscape, photo compressor, image size and many other compression applications. Web-based applications should be able to automatically resize the file according to the appropriate file size so that it can save time and speed up the process. To deal with the problem of automatic compression, the user must provide or build a web-based application that can directly compress files on the web because automatic file or image compression is available on the website without compressing them separately beforehand. With this automatic compression form application, it certainly reduces costs. (quota) internet, shortening time and also making it easier for users or users when uploading files on the web. The algorithm that is suitable for resizing files in online web applications is the hufman algorithm.

F-Divergences and Their Applications in Lossy Compression and Bounding Generalization Error

In this paper, we provide three applications for f-divergences: (i) we introduce Sanov’s upper bound on the tail probability of the sum of independent random variables based on super-modular f-divergence and show that our generalized Sanov’s bound strictly improves over ordinary one, (ii) we consider the lossy compression problem which studies the set of achievable rates for a given distortion and code length. We extend the rate-distortion function using mutual f-information and provide new and strictly better bounds on achievable rates in the finite blocklength regime using super-modular f-divergences, and (iii) we provide a connection between the generalization error of algorithms with bounded input/output mutual f-information and a generalized rate-distortion problem. This connection allows us to bound the generalization error of learning algorithms using lower bounds on the f-rate-distortion function. Our bound is based on a new lower bound on the rate-distortion function that (for some examples) strictly improves over previously best-known bounds.

An approximation algorithm for high-dimensional table compression on balanced [formula omitted]-partite graph

This paper considers the high-dimensional table compression problem on balanced k-partite graph. The objective is to choose half of the vertices from each of the k-partite to maximize the total weight of edges connecting the chosen vertices. Our main contribution is a 0.8785-approximation algorithm for the high-dimensional table compression problem by introducing the α-independent solutions through Lasserre semidefinite programming. This new algorithm improved two previous low-dimensional results, namely the 0.8731-approximation algorithm of Wu et al. for the one-dimensional case and the 0.6708-approximation of Xu and Du for the two-dimensional case.

Information aggregation and data compression based on PLC for distributed photovoltaic integration

The massive integration of distributed photovoltaics (PVs) into power grid causes rapid growth of data collection, which has put significant pressure on power line communication (PLC). Edge computing enables efficient data compression on edge-side gateway to reduce data size. However, it still faces challenges such as the coupling between transmission delay and compression delay, as well as the coupling between data compression ratio selection and gateway selection. In this paper, we formulate the information aggregation and data compression problem to minimize the total delay, and develop a delay-aware upper confidence bound (UCB) algorithm. It sets the negative value of delay as the learning reward to achieve delay-aware learning, and jointly optimizes gateway and data compression ratio selection based on upper confidence bound ranking and Karush-Kuhn-Tucker (KKT) conditions. Simulation results demonstrate the superior performance of the proposed algorithm.

A discretization‐convergent level‐set‐discrete‐element‐method using a continuum‐based contact formulation

AbstractThe level‐set‐discrete‐element‐method (LS‐DEM) was developed to overcome the shape limitation of traditional discrete element method. LS‐DEM's shape generality relies on a node‐based surface discretization of grain boundary, and it has been used to shed new light of a variety of granular mechanics applications with realistically shaped grains and structural assemblies made of unbonded building blocks. Due to the node‐based discretization of grain boundary, the original LS‐DEM is discretization‐sensitive and it suffers from divergence of the response with discretization refinement, particularly for highly compressible problems. Previous studies have identified and addressed this issue in different ways, each with its own advantages and shortcomings. Here, we propose a methodologically‐rigorous and computationally‐efficient adapted formulation which solves LS‐DEM's discretization‐sensitivity issue. It adopts the classical contact description of continuum mechanics, wherein the contact interactions are traction‐based. We demonstrate the convergence of the adapted LS‐DEM in several highly compressible cases studies, show that it is key to correctly capturing the mechanical response, and compare it to alternative formulations.

Structural response of steel sandwich panels with PUR foam core subjected to edgewise compression: Experimental, numerical, and analytical considerations

With the objective of leading to innovative lightweight structural components in the construction sector using sandwich panels with polyurethane foam and steel face sheets, different studies were conducted on such structures, including experimental, numerical, and analytical campaigns. The experimental campaign includes the mechanical characterisation of the constituent materials and the edgewise compressive test of small-scale sandwich panels. The test protocols are detailed, and their influence on the results is assessed. Through the use of digital image correlation recordings, deformation patterns and failure mechanisms in the constituent materials and the composite structure are identified. Material constitutive models are calibrated based on the experimentally obtained results. The accuracy of such models is first evaluated against the stress–strain curves recorded during the experiments. In the second stage, the response of the sandwich panels subjected to edgewise compressive loading is simulated. A parametric analysis on the sensitivity of the numerical models to the initial geometrical imperfections is shown to be fundamental to capturing the behaviour of the sandwich panels. Furthermore, the experimental and numerical results are compared to analytical solutions of the problem of the edgewise compression of sandwich panels. Finally, a thorough discussion of the obtained experimental, numerical and analytical results is presented, along with comparisons to results available in the literature.

Development of an implicit high-order Flux Reconstruction solver for high-speed flows on simplex elements

In this study, a significant emphasis is placed on the extension of high-order Flux Reconstruction (FR) schemes to cater to high-speed flows, specifically those involving high Mach numbers. The paper delves into the development and implementation of these FR schemes on both straight and curved-edged 2D and 3D simplex elements within the open-source COOLFluiD (Computational Object-Oriented Libraries for Fluid Dynamics) platform. The proposed FR solver provides more accurate detection of complex flow features over relatively coarser meshes when compared to their low-order peers (particularly in presence of shock waves), is fully implicit and able to simulate compressible flow problems governed by the Euler or Navier-Stokes equations on triangular and tetrahedral meshes. In order to simulate cases with shocks, a shock-capturing scheme previously developed for quadrilateral elements, was extended to tackle supersonic and hypersonic simulations. Extensive verification of the resulting FR solver (up to 7th order, a.k.a P6, for the solution polynomial and 3rd order a.k.a Q2 for the geometrical representation) has been performed on benchmark test cases with flow speeds ranging from subsonic to hypersonic, up to Mach 9.6. The obtained results have been favorably compared to those available in the literature.

Model Compression Algorithm via Reinforcement Learning and Knowledge Distillation

Traditional model compression techniques are dependent on handcrafted features and require domain experts, with a tradeoff between model size, speed, and accuracy. This study proposes a new approach toward resolving model compression problems. Our approach combines reinforcement-learning-based automated pruning and knowledge distillation to improve the pruning of unimportant network layers and the efficiency of the compression process. We introduce a new state quantity that controls the size of the reward and an attention mechanism that reinforces useful features and attenuates useless features to enhance the effects of other features. The experimental results show that the proposed model is superior to other advanced pruning methods in terms of the computation time and accuracy on CIFAR-100 and ImageNet dataset, where the accuracy is approximately 3% higher than that of similar methods with shorter computation times.

A cell-centred Eulerian volume-of-fluid method for compressible multi-material flows

We present a practical cell-centred volume-of-fluid method developed within a pure Eulerian setting for the simulation of compressible solid-fluid problems. The method builds on a previously published diffuse-interface Godunov-type scheme through the addition of a specialised mixed-cell update that is capable of maintaining sharp interfaces indefinitely. The mixed-cell update is local and may be viewed as an interface-sharpening extension to the underlying diffuse-interface scheme serving the same purpose as other techniques such as Tangent of Hyperbola INterface Capturing (THINC), and hence the method can be straightforwardly extended to include other coupled physics. We validate the method on a range of challenging test problems including a collapsing metal shell, cylinder impacts and the three-dimensional simulation of a buried explosive charge. Finally we demonstrate the robustness of the method, and its use in a multi-physics context, by modelling the BRL 105 mm unconfined shaped charge with reactive high-explosive burn and rate-sensitive plasticity.

Steady translating hollow vortex pair in weakly compressible flow

The weakly compressible translating hollow vortex pair is examined using the Imai–Lamla formula and a direct conformal mapping approach applied to a Rayleigh–Jansen expansion in the Mach number, M, taken to be small. The incompressible limit has been studied by Crowdy et al. (Eur. J. Mech. B Fluids 37 (2013), 180–186) who found explicit formulas for the solutions using conformal mapping. We improve upon their results by finding an explicit formula for the conformal map which had been left as an integral in Crowdy et al. (2013). The weakly compressible problem requires the solution of two boundary value problems in an annulus. This results in a linear parameter problem for the perturbed propagation velocity and speed along the vortex boundary. We find that two additional constraints are required to solve for the perturbed parameters. We require that the perturbation in vortex area and the perturbation in centroid separation to vanish. Three possible centroid definitions are given and the results worked out for each case. It is found that the correction to the propagation velocity is always negative, so that the vortex always slows down at first order due to compressible effects. Numerical results are consistent in the limit of small vortex size with the previous result by Leppington (J. Fluid Mech. 559 (2006), 45–55) that the propagation parameter is unchanged to O(M2).

A review on compaction and loosening of agricultural soils

Many soil experts and farmers are very concerned about the hard soil layers of agricultural soils. A layer known as hard soil is created when soil is compacted repeatedly over a long period of time. A type of physical soil degradation known as soil hardening modifies soil structure, restricts air and water penetration, and lessens root encroachment. These effects on the crust of the earth are still not fully understood. Therefore, to solve the worldwide dilemma of future food security, a thorough understanding of the subsurface processes is required. Compaction research must also be taken seriously because climatic conditions are shifting as a result of climate change, opening the door for the creation of novel mechanical tillers that can address compression problems. As a result, the focus of this study is on investigating the type, traits, causes, and effects of soil compaction on agricultural production as well as potential remedies using soil tillage.

Surrogate Lagrangian Relaxation: A Path to Retrain-Free Deep Neural Network Pruning

Network pruning is a widely used technique to reduce computation cost and model size for deep neural networks. However, the typical three-stage pipeline (i.e., training, pruning, and retraining (fine-tuning)) significantly increases the overall training time. In this article, we develop a systematic weight-pruning optimization approach based on surrogate Lagrangian relaxation (SLR), which is tailored to overcome difficulties caused by the discrete nature of the weight-pruning problem. We further prove that our method ensures fast convergence of the model compression problem, and the convergence of the SLR is accelerated by using quadratic penalties. Model parameters obtained by SLR during the training phase are much closer to their optimal values as compared to those obtained by other state-of-the-art methods. We evaluate our method on image classification tasks using CIFAR-10 and ImageNet with state-of-the-art multi-layer perceptron based networks such as MLP-Mixer; attention-based networks such as Swin Transformer; and convolutional neural network based models such as VGG-16, ResNet-18, ResNet-50, ResNet-110, and MobileNetV2. We also evaluate object detection and segmentation tasks on COCO, the KITTI benchmark, and the TuSimple lane detection dataset using a variety of models. Experimental results demonstrate that our SLR-based weight-pruning optimization approach achieves a higher compression rate than state-of-the-art methods under the same accuracy requirement and also can achieve higher accuracy under the same compression rate requirement. Under classification tasks, our SLR approach converges to the desired accuracy × faster on both of the datasets. Under object detection and segmentation tasks, SLR also converges 2× faster to the desired accuracy. Further, our SLR achieves high model accuracy even at the hardpruning stage without retraining, which reduces the traditional three-stage pruning into a two-stage process. Given a limited budget of retraining epochs, our approach quickly recovers the model’s accuracy.

Vector Quantized Variational Autoencoder-Based Compressive Sampling Method for Time Series in Structural Health Monitoring

The theory of compressive sampling (CS) has revolutionized data compression technology by capitalizing on the inherent sparsity of a signal to enable signal recovery from significantly far fewer samples than what is required by the Nyquist–Shannon sampling theorem. Recent advancement in deep generative models, which can represent high-dimension data in a low-dimension latent space efficiently when trained with big data, has been used to further reduce the sample size for image data compressive sampling. However, compressive sampling for 1D time series data has not significantly benefited from this technological progress. In this study, we investigate the application of different architectures of deep neural networks suitable for time series data compression and propose an efficient method to solve the compressive sampling problem on one-dimensional (1D) structural health monitoring (SHM) data, based on block CS and the vector quantized–variational autoencoder model with a naïve multitask paradigm (VQ-VAE-M). The proposed method utilizes VQ-VAE-M to learn the data characteristics of the signal, replaces the “hard constraint” of sparsity to realize the compressive sampling signal reconstruction and thereby does not need to select the appropriate sparse basis for the signal. A comparative analysis against various CS methods and other deep neural network models was performed in both synthetic data and real-world data from two real bridges in China. The results have demonstrated the superiority of the proposed method, with achieving the smallest reconstruction error of 0.038, 0.034 and 0.021, and the highest reconstruction accuracy of 0.882, 0.892 and 0.936 for compression ratios of 4.0, 2.66, and 2.0, respectively.

An adaptive, training‐free reduced‐order model for convection‐dominated problems based on hybrid snapshots

AbstractThe vast majority of reduced‐order models (ROMs) first obtain a low dimensional representation of the problem from high‐dimensional model (HDM) training data which is afterwards used to obtain a system of reduced complexity. Unfortunately, convection‐dominated problems generally have a slowly decaying Kolmogorov ‐width, which makes obtaining an accurate ROM built solely from training data very challenging. The accuracy of a ROM can be improved through enrichment with HDM solutions; however, due to the large computational expense of HDM evaluations for complex problems, they can only be used parsimoniously to obtain relevant computational savings. In this work, we exploit the local spatial coherence often exhibited by these problems to derive an accurate, cost‐efficient approach that repeatedly combines HDM and ROM evaluations without a separate training phase. Our approach obtains solutions at a given time step by either fully solving the HDM or by combining partial HDM and ROM solves. A dynamic sampling procedure identifies regions that require the HDM solution for global accuracy and the reminder of the flow is reconstructed using the ROM. Moreover, solutions combining both HDM and ROM solves use spatial filtering to eliminate potential spurious oscillations that may develop. We test the proposed method on inviscid compressible flow problems and demonstrate speedups up to a factor of five.