Hybrid Compression Method Research Articles

Enhancing IoT Data Integrity and Effectiveness through hybrid Compression Method: A Step Towards Energy Efficiency

The expansion of the Internet of Things (IoT) has magnified the challenge of managing data generated by IoT devices, notably in meteorological applications like temperature and humidity monitoring. This research addresses the imperative of efficiently reducing IoT data volume while preserving data integrity and underscores the significant implications for energy consumption. Our approach involved a two-fold strategy, employing the DHT11 sensor and ESP32 microcontroller for data collection, followed by an exploration of various data compression algorithms: delta encoding, run-length encoding (RLE), variable-length integer encoding (VLI), and bit-packing. The strategic combination of RLE and delta encoding yielded an exceptional compression rate of 98%. Beyond data reduction, this methodology offers energy savings by minimizing data transmission times, evidenced by the swift 133-microsecond compression process. Furthermore, the seamless transmission of compressed IoT data to Azure Cloud not only reduced cloud storage costs but also optimized storage space, contributing to energy efficiency. This research illuminates the significance of data compression in mitigating the environmental impact of IoT technologies, fostering a greener, more energy-conscious future.

HMC: Hybrid model compression method based on layer sensitivity grouping.

Previous studies have shown that deep models are often over-parameterized, and this parameter redundancy makes deep compression possible. The redundancy of model weight is often manifested as low rank and sparsity. Ignoring any part of the two or the different distributions of these two characteristics in the model will lead to low accuracy and a low compression rate of deep compression. To make full use of the difference between low-rank and sparsity, a unified framework combining low-rank tensor decomposition and structured pruning is proposed: a hybrid model compression method based on sensitivity grouping (HMC). This framework unifies the existing additive hybrid compression method (AHC) and the non-additive hybrid compression method (NaHC) proposed by us into one model. The latter group the network according to the sensitivity difference of the convolutional layer to different compression methods, which can better integrate the low rank and sparsity of the model compared with the former. Experiments show that our approach achieves a better trade-off between test accuracy and compression ratio when compressing the ResNet family of models than other recent compression methods using a single strategy or additive hybrid compression.

A hybrid stochastic interpolation and compression method for kernel matrices

Kernel functions play a pivotal role in a wide range of scientific computing and machine learning problems, but they ofter result in dense kernel matrices that impose great challenges in computational costs at large scale. To address this issue, we develop a set of fast kernel matrix compressing algorithms, which can reduce computation cost of matrix operations in the related applications. The foundation of these algorithms is the polyharmonic spline interpolation, which encompass a set of radial basis functions that allow flexible choices of interpolating nodes, and a set of polynomial basis functions that guarantee the solvability and convergence of the interpolation. With these properties, original data points in the interacting kernel function can be randomly sampled with great flexibility, so the proposed method is suitable for complicated data structures, such as high-dimensionality, random distribution, or manifold. To further boost the algorithm accuracy and efficiency, our scheme incorporates a QR sampling strategy, and combined with a recently developed fast stochastic SVD to form a hybrid method. If the overall number of degree of freedom is N, then the compressing algorithm has complexity of O(N) for low-rank matrices, and O(Nlog⁡N) for general matrices with a hierarchical structure. Numerical results for data on various domains and different kernel functions validate the accuracy and efficiency of the proposed method.

Deep Hybrid Compression Network for Lidar Point Cloud Classification and Segmentation

Extensive research on deep neural networks for LiDAR point clouds has contributed inexhaustible momentum to the development of computer 3D vision applications. However, storage and energy consumption have always been a challenge for deploying these deep models on mobile devices. Quantization provides a feasible route, of which current primary research is focused on uniform bit-width quantization without considering different layers or filters’ sensitivity to different bit-widths. This article proposes a novel hybrid compression method based on relaxed mixed-precision quantization, relaxed weights pruning, and knowledge distillation to overcome the limitations of uniform quantization illustrated above, while further improving model accuracy and reducing model memory consumption. It employs a differentiable searching method to search for the optimal bit allocation and weight sparsity, while conducting feature distillation, accordingly considering the feature degradation by pooling operation in point cloud deep models. The proposed method combines three compression techniques to balance the trade-off between compression accuracy and model size. Pruning alleviates the increasing memory consumption problem caused by mixed-precision quantization, while distillation improves compression accuracy without increasing model size. The experiments validate that the proposed method outperforms state-of-the-art typical uniform quantization methods in terms of accuracy with an acceptable and relatively competitive compression performance.

PHY: A performance-driven hybrid communication compression method for distributed training

Distributed training is needed to shorten the training time of deep neural networks. However, the communication overhead often hurts performance efficiency, especially in a distributed computing environment with limited network bandwidth. Hence, gradient compression techniques have been proposed to reduce communication time. But, compression also has the risk of causing lower model accuracy and longer training time due to compression loss and compression time. As a result, compression may not consistently achieve desired results, and there are limited discussions on when and which compression should be used. To address this problem, we propose a performance-driven hybrid compression solution. We make three main contributions. (1) We describe a hybrid compression strategy that chooses the compression method for individual model gradients. (2) We build an offline performance estimator and an online loss monitor to ensure the compression decision can minimize training time without sacrificing mode accuracy. (3) Our implementation can be imported to existing deep learning frameworks and applicable to a wide range of compression methods. Up to 3.6x training performance speedup was observed compared to other state-of-the-art methods.

A Hybrid Compression Method for Compound Power Quality Disturbance Signals in Active Distribution Networks

A Hybrid Compression Method for Compound Power Quality Disturbance Signals in Active Distribution Networks

Hybrid Compression Optimization Based Rapid Detection Method for Non-Coal Conveying Foreign Objects.

The existence of conveyor foreign objects poses a serious threat to the service life of conveyor belts, which will cause abnormal damage or even tearing, so fast and effective detection of conveyor foreign objects is of great significance to ensure the safe and efficient operation of belt conveyors. Considering the need for the foreign object detection algorithm to operate in edge computing devices, this paper proposes a hybrid compression method that integrates network sparse, structured pruning, and knowledge distillation to compress the network parameters and calculations. Combined with a Yolov5 network for practice, three structured pruning strategies are specifically proposed, all of which are proven to have achieved a good compression effect. The experiment results show that under the pruning rate of 0.9, the proposed three pruning strategies can achieve more than 95% compression for network parameters, more than 90% compression for the computation, and more than 90% compression for the size of the network model, and the optimized network is able to accelerate inference on both Central Processing Unit (CPU) and Graphic Processing Unit (GPU) hardware platforms, with a maximum speedup of 70.3% on the GPU platform and 157.5% on the CPU platform, providing an excellent real-time performance but also causing a large accuracy loss. In contrast, the proposed method balances better real-time performance and detection accuracy (>88.2%) when the pruning rate is at 0.6~0.9. Further, to avoid the influence of motion blur, a method of introducing prior knowledge is proposed to improve the resistance of the network, thus strongly ensuring the detection effect. All the technical solutions proposed are of great significance in promoting the intelligent development of coal mine equipment, ensuring the safe and efficient operation of belt conveyors, and promoting sustainable development.

A hybrid compression method for the NMR data based on window averaging and Discrete Cosine Transform

A large amount of raw echo data is collected by NMR logging, especially in multi-dimensional NMR logging, and a direct inversion for the raw echo data is time-consuming, so the echo data collected need to be compressed before inversion to improve inversion speed. This paper proposes a multi-dimensional NMR data compression method based on Discrete Cosine Transform (DCT). Utilizing the energy concentration characteristics of DCT, the low-frequency information of the echo data is retained to realize the compression of raw echo data. On this basis, a hybrid compression (WAC-DCT) method is proposed, which uses the window averaging method (WAC) to pre-compress the raw echo data, and then uses the DCT compression method to compress the pre-compressed echo data again to achieve a high compression rate. The compression effects of the WAC method, the FastSVD method, the WAC-SVD method, the DCT method and the WAC-DCT method are compared by numerical simulation. The results show that some fluid information is lost in the DCT compression process with a low compression rate. The FastSVD method, the WAC-SVD method and the WAC-DCT method all have the characteristic of a high compression rate. When the number of compressed echoes is very small, the WAC-DCT method has obvious advantage in T1 distribution. Compared with the case of without compression, the WAC-DCT method not only ensures a high compression rate, but also effectively reduces the inversion time, which has a good application effect in multi-dimensional NMR data compression.

A Hybrid Compression Method for Medical Images Based on Region of Interest Using Artificial Neural Networks

The number of medical images being stored and communicated daily is rapidly increasing, according to the need for these images in medical diagnoses. Hence, the storage space and bandwidths required to store and communicate these images are exponentially increasing, which has brought attention toward compressing these images. In this study, a new compression method is proposed for medical images based on convolutional neural networks. The proposed neural network consists of two main stages: a segmentation stage and an autoencoder. The segmentation stage is used to recognize the Region of Interest (ROI) in the image and provide it to the autoencoder stage, so more emphasis on the information of the ROI is applied. The autoencoder part of the neural network contains a bottleneck layer that has one-eighth of the dimensions of the input image. The values in this layer are used to represent the image, while the following layers are used to decompress the images, after training the neural network. The proposed method is evaluated using the CLEF MED 2009 dataset, where the evaluation results show that the method has significantly better performance, compared to the existing state-of-the-art methods, by providing more visually similar images using less data.

Luminance approximated vector quantization algorithm to retain better image quality of the decompressed image

Some compressed images using Vector Quantization algorithm suffers from blocking artifacts which degrades the visual appeal of the image. Present study proposes a hybrid vector quantization method applicable on de-correlated color model. As luminance channel carries image information and loss of image information results in degradation of the visual appeal of an image, so aim of this study is focused on retaining more image information during compression process. For luminance channel compression, a new four level quantization based compression method is developed. Luminance channel is partitioned into smaller blocks. Then for each block, four level quantization is applied which local to the current block only. This results many level luminance value effectively for the whole image. It helps to retain better information. Chrominance channels are compressed using conventional Vector Quantization. This hybrid compression method improves visual quality of the decompressed image reasonably compared to VQ. The proposed method is applied on many standard images found in literature and images of UCIDv.2 color image database. Results are analyzed in terms of Peak Signal to Noise Ratio, Structure Similarity Index and space requirement reduction for compressed image using the method. Experimental results show that proposed method retains better quality of image in terms of PSNR and SSIM than Vector Quantization and Modified Vector Quantization. This method reduces storage space requirement for the compressed images in the range of 84% to 89%.

Low Power FPGA Implementation of Multi-View Video Coding with Hybrid Compression and Decompression Algorithm

The video is one of the most useful and most appealing medium to represent some information. More usage of digital multi-media via communications media, wireless communications, intranet, internet and cellular mobile leads to the uncontrollable growth of data in media. The video compression technique is used in this research work to improve the processing speed of the entire system. In this work, Low Cost - Multi Video Coding - Hybrid Compression and Decompression (LC-MVC-HCD) method is used to reduce computation complexity. The combinational of Discrete Wavelet Transform (DWT) and Discrete Cosine Transform (DCT) algorithms are denoted as hybrid algorithm. Based on this hybrid algorithm, the compression process is performing which improves the video coding efficiency of MVC. The LC-MVC-HCD methodwas implemented in the Matlab, Xilinx and Cadence tool. In Application Specific Integrated Circuit (ASIC) implementation, the area, power, and delay minimized by using the cadence encounter tool with 180nm and 45nm technology. In Field Programmable Gate Array (FPGA) implementation, the number of Lookup Tables (LUTs), slice, and flip-flop are minimized based on two different kinds of Virtex devices such as Virtex -6 and Virtex-7. In Matlab, Peak Signal to Noise Ratio (PSNR), computational time and bit error rate were analyzed for the LC-MVC-HCD method. The experimental outcome showed that the proposed methodology has improved performance ASIC and FPGA up to 2-3% compared to existing methods like Direct mode decision MVC and LC-MVC-DWT.

Hybrid image compression technique using oscillation concept & quasi fractal

In medical images, especially for brain images ROI is very important for diagnosis. ROI is very important compare to other portion of an image. Here ROI is included in hybrid coding algorithm for effective image compression. Compression method gives better results using hybrid algorithm. In this paper, we have used hybrid compression method, Lossless used for ROI portion and for non-ROI portion the lossy compression techniques has been used. The experimental results shows that better Compression Ratio (CR) with acceptable PSNR has been achieved using hybrid technique based on Morphological band pass filter and Adaptive thresholding for ROI.

IMPLEMENTASI DAN ANALISIS KOMPRESI HYBRID PADA CITRA MEDIS DIGITAL HASIL RONTGEN KANKER PAYUDARA

Digital image compression is an attempt to transform the data or symbols of composing digital images into data or other symbols without causing significant changes in the digital image to the human views. The goal is to reduce the redundancy of the data contained in the image so it can be stored or transmitted efficiently. Breast cancer is one of the greatest cancer cases in Indonesia which requires so many digital data x-rays. This research is intended to compress digital X-ray images of breast cancer with more efficient size without altering or reducing its information and it still can be assessed by human view. Using hybrid compression method, they are 2D DCT (Two Dimensional Discret Cosine Transform) and Huffman Method, Medical expert gave 3 of 4 rate when it reach 34.36% compression ratio with 40dB PSNR. This test result used 0 till 3 depth of zero in 2D DCT quantization process.

Accelerating Recurrent Neural Networks: A Memory-Efficient Approach

Recurrent neural networks (RNNs) have achieved the state-of-the-art performance on various sequence learning tasks due to their powerful sequence modeling capability. However, RNNs usually require a large number of parameters and high computational complexity. Hence, it is quite challenging to implement complex RNNs on embedded devices with stringent memory and latency requirement. In this paper, we first present a novel hybrid compression method for a widely used RNN variant, long-short term memory (LSTM), to tackle these implementation challenges. By properly using circulant matrices, forward nonlinear function approximation, and efficient quantization schemes with a retrain-based training strategy, the proposed compression method can reduce more than 95% of memory usage with negligible accuracy loss when verified under language modeling and speech recognition tasks. An efficient scalable parallel hardware architecture is then proposed for the compressed LSTM. With an innovative chessboard division method for matrix-vector multiplications, the parallelism of the proposed hardware architecture can be freely chosen under certain latency requirement. Specifically, for the circulant matrix-vector multiplications employed in the compressed LSTM, the circulant matrices are judiciously reorganized to fit in with the chessboard division and minimize the number of memory accesses required for the matrix multiplications. The proposed architecture is modeled using register transfer language (RTL) and synthesized under the TSMC 90-nm CMOS technology. With 518.5-kB on-chip memory, we are able to process a 512×512 compressed LSTM in 1.71 μs, corresponding to 2.46 TOPS on the uncompressed one, at a cost of 30.77-mm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> chip area. The implementation results demonstrate that the proposed design can achieve significantly high flexibility and area efficiency, which satisfies many real-time applications on embedded devices. It is worth mentioning that the memory-efficient approach of accelerating LSTM developed in this paper is also applicable to other RNN variants.

Hybrid Compression of Hyperspectral Images Based on PCA With Pre-Encoding Discriminant Information

It has been shown that image compression based on principal component analysis (PCA) provides good compression efficiency for hyperspectral images. However, PCA might fail to capture all the discriminant information of hyperspectral images, since features that are important for classification tasks may not be high in signal energy. To deal with this problem, we propose a hybrid compression method for hyperspectral images with pre-encoding discriminant information. A feature extraction method is first applied to the original images, producing a set of feature vectors that are used to generate feature images and then residual images by subtracting the feature-reconstructed images from the original ones. Both feature images and residual images are compressed and transmitted. Experiments on data from the Airborne Visible/Infrared Imaging Spectrometer sensor indicate that the proposed method provides better compression efficiency with improved classification accuracy than conventional compression methods.

A hybrid method for segmented test data compression

This paper presents a hybrid compression method for scan-based testing. Four distinctive compression techniques, i.e. frequency-directed run-length codes (FDR), inverting FDR, option coding technique, and optimal selective Huffman code are selected for compression of a given test set. To lift the potential compression limit, we properly cluster scan cells to classify test data as four subsets respectively, by majorities of 0s, 1s, Xs, and a roughly equal number of 0s and 1s to match specific properties of the compression techniques. Segments are then, respectively, compressed by the technique that fits best to achieve an overall good compression effect. Experimental results for the larger ISCAS’89 benchmark circuits show that up to 69.9% of average compression ratio and significant test application time reduction can be achieved.

A hybrid compression method for head-related transfer functions

The paper proposes a hybrid compression method to resolve the storage problem of a large number of head-related transfer functions (HRTFs). First, each HRTF is approximated by a minimum-phase HRTF and an all pass filter whose group delay equals the interaural time delay (ITD). Second, principal component analysis is applied to the entire HRTF set to derive several basis functions, with a weight vector set defining the contribution of the basis functions to each HRTF. Third, the weight set is vector quantized with the designed codebook. At last, the ITD is curved surface fitted with a cosine series bivariate polynomial. As a result, the HRTF can be reconstructed from the basis functions, codebook indexes, and ITD polynomial coefficients. Simulation results reveal that the proposed method may reduce the data size greatly with similar reconstruction precision comparing with the principal component analysis method.

Wavelet-based hybrid ECG compression technique

In this paper, a new wavelet-based hybrid electrocardiogram (ECG) data compression technique is proposed. Firstly, in order to fully utilize the two correlations of heartbeat signals, 1-D ECG data are segmented and aligned to a 2-D data arrays. Secondly, 2-D wavelet transform is applied to the constructed 2-D data array. Thirdly, the set partitioning hierarchical trees (SPIHT) method and the vector quantization (VQ) method are modified, according to the individual characteristic of different coefficient subband and the similarity between the subbands. Finally, a hybrid compression method of the modified SPIHT and VQ is employed to the wavelet coefficients. Records selected from the MIT/BIH arrhythmia database are tested. The experimental results show that the proposed method is suitable for various morphologies of ECG data, and that it achieves high compression ratio with the characteristic features well preserved.