Compression Algorithm Research Articles

Cerebrospinal fluid metabolomes of treatment-resistant depression subtypes and ketamine response: a pilot study.

Depression is a disorder with variable presentation. Selecting treatments and dose-finding is, therefore, challenging and time-consuming. In addition, novel antidepressants such as ketamine have sparse optimization evidence. Insights obtained from metabolomics may improve the management of patients. The objective of this study was to determine whether compounds in the cerebrospinal fluid (CSF) metabolome correlate with scores on questionnaires and response to medication. We performed a retrospective pilot study to evaluate phenotypic and metabolomic variability in patients with treatment-resistant depression using multivariate data compression algorithms. Twenty-nine patients with treatment-resistant depression provided fasting CSF samples. Over 300 metabolites were analyzed in these samples with liquid chromatography-mass spectrometry. Chart review provided basic demographic information, clinical status with self-reported questionnaires, and response to medication. Of the 300 metabolites analyzed, 151 were present in all CSF samples and used in the analyses. Hypothesis-free multivariate analysis compressed the resultant data set into two dimensions using Principal Component (PC) analysis, accounting for ~ 32% of the variance. PC1 accounted for 16.9% of the variance and strongly correlated with age in one direction and 5-methyltetrahydrofolate, homocarnosine, and depression and anxiety scores in the opposite direction. PC2 accounted for 15.4% of the variance, with one end strongly correlated with autism scores, male gender, and cognitive fatigue scores, and the other end with bipolar diagnosis, lithium use, and ethylmalonate disturbance. This small pilot study suggests that complex treatment-resistant depression can be mapped onto a 2-dimensional pathophysiological domain. The results may have implications for treatment selection for depression subtypes.

Impact of Audio Data Compression on Feature Extraction for Vocal Biomarker Detection: Validation Study.

Vocal biomarkers, derived from acoustic analysis of vocal characteristics, offer noninvasive avenues for medical screening, diagnostics, and monitoring. Previous research demonstrated the feasibility of predicting type 2 diabetes mellitus through acoustic analysis of smartphone-recorded speech. Building upon this work, this study explores the impact of audio data compression on acoustic vocal biomarker development, which is critical for broader applicability in health care. The objective of this research is to analyze how common audio compression algorithms (MP3, M4A, and WMA) applied by 3 different conversion tools at 2 bitrates affect features crucial for vocal biomarker detection. The impact of audio data compression on acoustic vocal biomarker development was investigated using uncompressed voice samples converted into MP3, M4A, and WMA formats at 2 bitrates (320 and 128 kbps) with MediaHuman (MH) Audio Converter, WonderShare (WS) UniConverter, and Fast Forward Moving Picture Experts Group (FFmpeg). The data set comprised recordings from 505 participants, totaling 17,298 audio files, collected using a smartphone. Participants recorded a fixed English sentence up to 6 times daily for up to 14 days. Feature extraction, including pitch, jitter, intensity, and Mel-frequency cepstral coefficients (MFCCs), was conducted using Python and Parselmouth. The Wilcoxon signed rank test and the Bonferroni correction for multiple comparisons were used for statistical analysis. In this study, 36,970 audio files were initially recorded from 505 participants, with 17,298 recordings meeting the fixed sentence criteria after screening. Differences between the audio conversion software, MH, WS, and FFmpeg, were notable, impacting compression outcomes such as constant or variable bitrates. Analysis encompassed diverse data compression formats and a wide array of voice features and MFCCs. Wilcoxon signed rank tests yielded P values, with those below the Bonferroni-corrected significance level indicating significant alterations due to compression. The results indicated feature-specific impacts of compression across formats and bitrates. MH-converted files exhibited greater resilience compared to WS-converted files. Bitrate also influenced feature stability, with 38 cases affected uniquely by a single bitrate. Notably, voice features showed greater stability than MFCCs across conversion methods. Compression effects were found to be feature specific, with MH and FFmpeg showing greater resilience. Some features were consistently affected, emphasizing the importance of understanding feature resilience for diagnostic applications. Considering the implementation of vocal biomarkers in health care, finding features that remain consistent through compression for data storage or transmission purposes is valuable. Focused on specific features and formats, future research could broaden the scope to include diverse features, real-time compression algorithms, and various recording methods. This study enhances our understanding of audio compression's influence on voice features and MFCCs, providing insights for developing applications across fields. The research underscores the significance of feature stability in working with compressed audio data, laying a foundation for informed voice data use in evolving technological landscapes.

A Comparison of Machine Learning-Based and Conventional Technologies for Video Compression

The growing demand for high-quality video transmission over bandwidth-constrained networks and the increasing availability of video content have led to the need for efficient storage and distribution of large video files. To improve the latter, this article offers a comparison of six video compression methods without loss of quality. Particularly, H.255, VP9, AV1, convolutional neural network (CNN), recurrent neural network (RNN), and deep autoencoder (DAE). The proposed decision is to use a dataset of high-quality videos to implement and compare the performance of classical compression algorithms and algorithms based on machine learning. Evaluations of the compression efficiency and the quality of the received images were made on the basis of two metrics: PSNR and SSIM. This comparison revealed the strengths and weaknesses of each approach and provided insights into how machine learning algorithms can be optimized in future research. In general, it contributed to the development of more efficient and effective video compression algorithms that can be useful for a wide range of applications.

Literature Review on Digital Image Processing and Its Techniques

Purpose: Understand the current state of knowledge in the field of digital image processing. Identify key concepts, theories, and methodologies that have been explored in previous research. Explore the various techniques and algorithms used in digital image processing. Evaluate the strengths and weaknesses of existing methods. Identify gaps or limitations in current techniques. Provide a context for the research by summarizing and synthesizing relevant studies. Show how different studies contribute to the overall understanding of digital image processing. Identify gaps in the existing literature that your research can address. Determine areas where further investigation is needed. Offer a foundation for discussion and interpretation of your results in the context of existing literature. Enable you to relate your findings to the broader field of digital image processing. Design/Methodology/Approach: Clearly define the scope of your literature review (e.g., specific techniques, applications, or periods). State the objectives of the literature review, such as identifying trends, evaluating methodologies, or addressing research gaps. Develop a comprehensive search strategy to identify relevant literature. Utilize academic databases, journals, conferences, and other reputable sources. Use a combination of keywords, Boolean operators, and controlled vocabulary (e.g., MeSH terms) to refine search queries. Systematically review and select relevant literature based on the established criteria. Document the process, including databases searched, keywords used, and reasons for inclusion/exclusion. Identify gaps in the existing literature and propose potential avenues for future research. Discuss the implications of these gaps for the advancement of knowledge in the field Findings/Results: Literature often discusses various image enhancement techniques such as histogram equalization, contrast stretching, and spatial filtering. Researchers explore the effectiveness of these techniques in improving image quality for different applications. Segmentation methods, including thresholding, region-based segmentation, and clustering algorithms, are frequently discussed. Object recognition and classification techniques using features like texture, color, and shape are common topics. Different image compression algorithms, such as JPEG, JPEG2000, and various wavelet-based methods, are often compared in terms of compression ratio and quality. The literature might address real-time image processing challenges and solutions, especially in applications like video surveillance, autonomous vehicles, and augmented reality. Originality/Value: Ensure that the literature review comprehensively covers key and recent works in digital image processing. This includes foundational theories, algorithms, and applications. Identify seminal papers, landmark studies, and recent advancements to create a timeline of the field's development. Paper Type: Review of existing literature

Lossless and Near-Lossless Compression Algorithms for Remotely Sensed Hyperspectral Images.

Rapid and continuous advancements in remote sensing technology have resulted in finer resolutions and higher acquisition rates of hyperspectral images (HSIs). These developments have triggered a need for new processing techniques brought about by the confined power and constrained hardware resources aboard satellites. This article proposes two novel lossless and near-lossless compression methods, employing our recent seed generation and quadrature-based square rooting algorithms, respectively. The main advantage of the former method lies in its acceptable complexity utilizing simple arithmetic operations, making it suitable for real-time onboard compression. In addition, this near-lossless compressor could be incorporated for hard-to-compress images offering a stabilized reduction at nearly 40% with a maximum relative error of 0.33 and a maximum absolute error of 30. Our results also show that a lossless compression performance, in terms of compression ratio, of up to 2.6 is achieved when testing with hyperspectral images from the Corpus dataset. Further, an improvement in the compression rate over the state-of-the-art k2-raster technique is realized for most of these HSIs by all four variations of our proposed lossless compression method. In particular, a data reduction enhancement of up to 29.89% is realized when comparing their respective geometric mean values.

A Reconstruction Algorithm for Compression Perception Based on Local Learning Dictionary and Sparsity Iterative Update

The global sparse representation method based on compressed sensing fails to capture the local texture and detail structure of an image. To address this, a local dictionary learning method based on the wavelet domain is proposed. The wavelet high-frequency subband sub-block classification and local dictionary learning are implemented using the FCM clustering method and K-L method, respectively. At the reconstruction end, a compressed sampling Matching pursuit algorithm based on iterative update of sparsity is proposed. This adaptive iteration can effectively reconstruct the original image under unknown sparsity conditions. Simulation experiments show that compared to existing reconstruction methods, this approach has the advantages of simple computation, strong adaptive sparse representation ability for the original image, and superior reconstructed image performance.

A Time-sensitive Traffic Scheduling Algorithm for Industrial Network of Power Grid

Industrial Internet communications in power scenarios are highly time-sensitive and require low network latency and high stability. Therefore, in order to ensure the smooth operation of the automated production line, we propose a time-sensitive traffic scheduling (TsTS) algorithm combined with SRv6 technology based on the application background of industrial and civil converged networks. The method is: through the previous hop node sensing of network congestion Link congestion, and then adjust the detour path for time-sensitive traffic through Dijkstra’s algorithm. Moreover, in order to effectively reduce the link overhead of SRH and prevent bandwidth waste caused by large IPv6 addresses, we also propose a node list compression algorithm for detour paths. The core method is to put only necessary nodes into SRH, that is, Sensitive traffic can be forwarded with the help of ordinary routing. Experiments show that this method performs well in terms of delay and link utilization.

Joint watermarking encryption compression algorithm for securing an e-healthcare application

Joint watermarking encryption compression algorithm for securing an e-healthcare application

Lossy compression of infrared videos in WEST (W Environment in Steady-state Tokamak).

During the 2019 C4 experimental campaign of the WEST (W Environment in Steady-state Tokamak) (France), the infrared diagnostic produced more than seven terabytes of uncompressed video data. Constraints on the computer infrastructure required for storage, backup, and especially offline access to infrared videos made the use of a compression algorithm mandatory. This paper proposes an innovative method to compress infrared videos with controlled temperature precision. This compression method is based on a controlled averaging of the video that maximizes the compression potential of standard lossless video codecs such as H264/AVC or HEVC. The combination of the loss introduction algorithm and the H264/AVC lossless video codec obtains the best compression ratio in the range of 8 to 41 with a maximum temperature error of 2 °C. This method also outperforms the JPEG-LS algorithm in terms of compression ratio and image quality for the same temperature precision.

A novel algorithm for ship characteristic points extraction based on density clustering

With the widespread application of ship Automatic Identification System (AIS) in maritime operations, a large number of ship trajectories become available. This study aims to improve the safety of ships navigating through densely trafficked areas and address the challenge of sufficient data exploration while fully describing the traffic conditions in these waters. To achieve this objective, traffic flow information is extracted from AIS data collected in Zhoushan waters. A combination of multi-algorithms is employed to extract the traffic flow frame, specifically, the Douglas-Peucker compression algorithm and trajectory intersection algorithm are utilised to identify the characteristic points of ship trajectories. Subsequently, a density clustering algorithm is applied to extract the three types of characteristic points: compressed trajectory points, intersection points, and ship position points, facilitating data mining efforts. The resulting initial traffic flow characteristic points are then subject to weighted fusion, followed by image superposition processing to create an overlapping map of the ship trajectories. This process culminates in the generation of a traffic flow frame for the region. The framework integrates various track characteristic points, offering insights into the distribution of essential routes in the vicinity waters, thereby providing a comprehensive depiction of ship traffic flow patterns. The proposed framework can be applied to the route planning and serve as a reference to the maritime authorities when selecting recommended shipping lanes.

An efficient bandwidth-adaptive gradient compression algorithm for distributed training of deep neural networks

In distributed deep learning with data parallelism, communication bottleneck throttles the efficiency of model training. Recent studies adopt versatile gradient compression techniques, with communication sparsification standing out as an effective approach for reducing the number of gradients to be transmitted. However, the deployment of gradient sparsification is adversely influenced by the change of network environment in real systems, and existing methods either neglect bandwidth dynamics during training or experience drastic fluctuation of compression ratios. In this paper, we propose ACE, a novel adaptive gradient compression mechanism with high communication efficiency under bandwidth variation. ACE adapts the sparsification ratio to the average bandwidth in a time window, other than following its dynamics exactly. To accurately compute the compression ratio, we first profile the compression time and model a single iteration time consisting of communication, computation and compression operations. We then present a practical model to fit the needed training rounds till convergence, and formulate an optimization problem to compute the optimal sparsification ratio. We conduct experiments on different DNN models in different network environments and compare various methods in terms of convergence and model quality. The experimental results show that ACE achieves up to 9.39× and 1.28× training speedups over fixed and state-of-the-art adaptive compression methods.

Through the compression glass: language complexity and the linguistic structure of compressed strings

Abstract Against the backdrop of the sociolinguistic-typological complexity debate which is all about measuring, comparing and explaining language complexity, this article investigates how Kolmogorov-based information theoretic complexity relates to linguistic structures. Specifically, the linguistic structure of text which has been compressed with the text compression algorithm gzip will be analysed. One implementation of Kolmogorov-based language complexity is the compression technique (Ehret, Katharina. 2021. An information-theoretic view on language complexity and register variation: Compressing naturalistic corpus data. Corpus Linguistics and Linguistic Theory (2). 383–410) which employs gzip to measure language complexity in naturalistic text samples. In order to determine what type of structures compression algorithms like gzip capture, and how these compressed strings relate to linguistically meaningful structures, gzip’s lexicon output is retrieved and subjected to an in-depth analysis. As a case study, the compression technique is applied to the English version of Lewis Carroll’s Alice’s Adventures in Wonderland and its lexicon output is extracted. The results show that gzip-like algorithms sometimes capture linguistically meaningful structures which coincide, for instance, with lexical words or suffixes. However, many compressed sequences are linguistically unintelligible or simply do not coincide with any linguistically meaningful structures. Compression algorithms like gzip thus crucially capture purely formal structural regularities. As a consequence, information theoretic complexity, in this context, is a linguistically agnostic, purely structural measure of regularity and redundancy in texts.

Comparative Analysis of Super-Resolution Algorithms for Image Compression

Comparative Analysis of Super-Resolution Algorithms for Image Compression

Statistical Analysis of the Negative–Positive Transformation in Image Encryption

The negative–positive transformation (NPT) is a widely employed technique for encrypting images on pixel blocks, commonly integrated into cryptosystems compatible with compression algorithms. The existing literature on NPT analysis can be categorized into two types: theoretical analyses with results that apply to any image, primarily focused on compression compatibility, and numerical analyses that report empirical results from specific images, some without explaining the causes of the security results, while others are only related to the compression performance. Consequently, there is a significant gap in understanding the implications of applying the NPT for data protection. For that reason, this paper conducts a theoretical statistical analysis, presenting, demonstrating, and verifying six theorems to understand the security contributions of NPT. Two theorems examine the shape of the image histogram and the scatter plot of adjacent pixels after the NPT application. The subsequent four theorems explore the influence of NPT on the mean, variance, covariance, and correlation within each pixel block. The findings indicate that the NPT generates images with symmetrical histograms, the correlation of pixel blocks remains invariant, and distinct vertical and horizontal reflections manifest on the scatter plot. These theorems are verified by encrypting the Lena image with four pixel-block sizes. The histogram symmetry passed the goodness-of-fit test at a significance level of 5%, revealing consistent results. The correlation of pixel blocks remained unchanged, and the scatter plot exhibited an x-shaped pattern. Therefore, as the NPT alone does not achieve desirable encryption results, such as uniform histograms, scatter plots, and decreasing correlation, cryptosystems should complement it with additional techniques.

Machine Learning-Based Kernel Selector for SpMV Optimization in Graph Analysis

Sparse Matrix and Vector multiplication (SpMV) is one of the core algorithms in various large-scale scientific computing and real-world applications. With the rapid development of AI and big data, the input vector in SpMV becomes sparse in many application fields. Especially in some graph analysis calculations, the sparsity of the input vector will change with the running of the program, and the non-zero element distribution of the adjacency matrix of some graph data has the power law property, leading to serious load imbalance, which requires additional optimization means. Therefore, the optimal SpMV kernel may be different, and a single SpMV kernel can no longer meet the acceleration requirements. In this paper, we propose a decision tree-based adaptive SpMV framework, named DTSpMV, that can automatically select appropriate SpMV kernels according to different input data in iterations of graph computation. Based on the analysis of computing patterns, bit-array compression algorithms, and serial and parallel algorithms, we encapsulate nine SpMV kernels within the framework. We explore machine learning-based kernel selectors in terms of both accuracy and runtime overhead. Experimental results on NVIDIA Tesla T4 GPU show that our adaptive framework achieves the arithmetic average performance improvement of \(152\% \) compared to the SpMV kernel in cuSPARSE.

TDLC: Tensor decomposition‐based direct learning‐compression algorithm for DNN model compression

SummaryAs a deep neural networks (DNNs) model compression method, learning‐compression (LC) algorithm based on pre‐trained models and matrix decomposition increases training time and ignores the structural information of models. In this manuscript, a tensor decomposition‐based direct LC (TDLC) algorithm without pre‐trained models is proposed. In TDLC, the pre‐trained model is eliminated, and tensor decomposition is first applied to LC algorithm to preserve the structural features of the model. There are two key steps in TDLC. An optimal rank selection method is first proposed in compression‐step (C‐step) of TDLC to find global optimal ranks of tensor decomposition. Second, TDLC utilizes cyclical learning rate, which is different from traditional monotonically learning rates schedule, to improve the generalization performance of uncompressed models in learning‐step (L‐step). TDLC obtains the optimal compression model by alternately optimizing L‐step and C‐step. TDLC is compared with 16 state‐of‐the‐art compression methods in experiments part. Extensive experimental results show that TDLC produces high‐accuracy compression models with high compression rate. Comparing with TDLC‐pre‐trained, TDLC notably achieves 30% training time shorten and 11% parameter reduction in Resnet32, while improving accuracy by 0.2%.

An investigation of machine learning-based video compression techniques

As video technology continues to seamlessly weave itself into the fabric of daily life, there is a growing need for enhanced storage and efficient video transmission. This surge in demand has led to heightened expectations and standards for video compression technology. Machine learning as an up-and-coming technology can play its advantages in the field of video compression. This article reviews the current state of research on combining video compression techniques with machine learning. The article provides an overview of various research avenues for enhancement, spanning from conventional video compression algorithms to the fusion of traditional compression frameworks with machine learning methodologies, and even the development of novel end-to-end compression algorithms. In additional, the article explores the possible various application scenarios of machine learning-based video compression algorithms based on the characteristics of such non-standard and arithmetic demanding algorithms. At the end, the article speculates on the future of video compression algorithms based on the content of the various studies reviewed in the article.

Implementasi Algoritma Arithmetic Coding Pada Aplikasi Kompresi File PDF

Currently, application users use information more and more commonly. However, there is a problem that is often encountered, namely the need for large data so that less storage space is needed. Apart from that, the small amount of storage space also affects the speed of sending data. One of the data that is often used is PDF format files. The solution to this problem is to compress the PDF file. Compression is a process of changing a collection of data into a form of code to save on data storage requirements so that the data used is not too large. The Arithemtic Coding algorithm is a data compression algorithm that can be used, by knowing the results of the compression ratio and redundancy process. In applying the Arithmetic Coding algorithm in this research, it can provide compression results which initially have a large size which can be compressed very well in PDF files and the compression results show a compression ratio of 2.15 while the Compression ratio is 46% and Redundancy 54% will be beneficial in sending and moving PDF files will be even easier

Efficient Framework to Manipulate Data Compression and Classification of Power Quality Disturbances for Distributed Power System

There is some risk of power quality disturbances at many stages of production, transformation, distribution, and energy consumption. The cornerstone for dealing with power quality problems is the characterization of power quality disturbances (PQDs). However, past research has focused on a narrow topic: noise disruption, overfitting, and training time. A new strategy is suggested to address this problem that combines efficient one-dimensional dataset compression with the convolutional neural network (CNN) classification algorithm. First, three types of compression algorithms: wavelet transform, autoencoder, and CNN, are proposed to be evaluated. According to the IEEE-1159 standard, the synthetic dataset was built with fourteen different PQD types. Furthermore, the PQD classification procedure integrated compressed data with the CNN classification algorithm. Finally, the suggested method demonstrates that combining CNN compression and classification methods can efficiently recognize PQDs. Even in noisy environments, PQD signal processing achieved up to 98.25% accuracy and managed the overfitting.

An autonomous wheelchair with health monitoring system based on Internet of Thing

Assistive powered wheelchairs will bring patients and elderly the ability of remain mobile without the direct intervention from caregivers. Vital signs from users can be collected and analyzed remotely to allow better disease prevention and proactive management of health and chronic conditions. This research proposes an autonomous wheelchair prototype system integrated with biophysical sensors based on Internet of Thing (IoT). A powered wheelchair system was developed with three biophysical sensors to collect, transmit and analysis users’ four vital signs to provide real-time feedback to users and clinicians. A user interface software embedded with the cloud artificial intelligence (AI) algorithms was developed for the data visualization and analysis. An improved data compression algorithm Minimalist, Adaptive and Streaming R-bit (O-MAS-R) was proposed to achieve a higher compression ratio with minimum 7.1%, maximum 45.25% compared with MAS algorithm during the data transmission. At the same time, the prototype wheelchair, accompanied with a smart-chair app, assimilates data from the onboard sensors and characteristics features within the surroundings in real-time to achieve the functions including obstruct laser scanning, autonomous localization, and point-to-point route planning and moving within a predefined area. In conclusion, the wheelchair prototype uses AI algorithms and navigation technology to help patients and elderly maintain their independent mobility and monitor their healthcare information in real-time.