The subject matter of the article is the process of lossy compression of multilook Synthetic Aperture Radar (SAR) images corrupted by multiplicative, spatially correlated speckle noise, with a focus on operation in the neighborhood of the potential Optimal Operation Point (OOP). The goal of the article is to analyze the existence and properties of the OOP for SAR image compression using the Better Portable Graphics (BPG) coder, and to develop a practical method for achieving compression near this point. The tasks to be solved are: to verify the existence of the OOP for simulated Sentinel-1-like SAR images according to both traditional peak signal-to-noise ratio (PSNR) and visual quality (PSNR-HVS-M) metrics; to investigate the relationship between the compression control parameter (Q) and the resulting image quality and compression ratio (CR); and to propose and describe a practical iterative procedure for determining the Q parameter value corresponding to the OOP without requiring access to the noise-free reference image. The methods used are: simulation of SAR images with speckle relative variance equal to 0.05 using noise-free Sentinel-2 data as a reference; lossy compression using the BPG coder with parameter Q varying from 1 to 51; quantitative assessment using PSNR and PSNR-HVS-M metrics; calculation of compression ratio; analysis of rate-distortion curves between different image pairs; statistical estimation of equivalent noise variance for input PSNR prediction. The following results were obtained: It has been demonstrated that an OOP exists for the BPG coder when compressing multilook SAR images, confirmed by both PSNR and PSNR-HVS-M metrics. The OOP provides PSNR and PSNR-HVS-M values several dB higher compared to the uncompressed noisy image while achieving very high compression ratios (CR > 180). The OOP was found at high Q values (Q=48-49), where the coder aggressively suppresses noise but also introduces content distortions. A key practical result is the proposed method for determining Q at the OOP. Conclusions. The scientific novelty of the obtained results is as follows: For the first time, the existence of the OOP has been comprehensively demonstrated for the BPG coder applied to multilook SAR images with realistic speckle properties, considering not only the standard PSNR but also the visual quality metric PSNR-HVS-M, although the OOP is less pronounced for the latter; a method for practical OOP approximation has been developed, which operates without the need for the original noise-free (true) image, relying instead on an estimation of the speckle noise power from the available noisy data, making it applicable in real-world SAR image processing and transmission scenarios.

With the rapid expansion of data lakes storing health data and hosting AI algorithms, a prominent concern arises: how safe is it to export machine learning models from these data lakes? In particular, deep network models, widely used for health data processing, encode information from their training dataset, potentially leading to the leakage of sensitive information upon export. This paper thoroughly examines this issue in the context of medical imaging data and introduces a novel data exfiltration attack based on image compression techniques.<br>This attack, termed Data Exfiltration by Compression, requires only access to a data lake and is based on lossless or lossy image compression methods.<br>Unlike previous data exfiltration attacks, it is compatible with any image processing task and depends solely on an exported network model without requiring any additional information collected during the training process. We explore various scenarios, and techniques to limit the size of the exported model and conceals the compression codes within the network.<br>Using two public datasets of CT and MR images, we demonstrate that this attack can effectively steal medical images and reconstruct them outside the data lake with high fidelity, achieving an optimal balance between compression and reconstruction quality. Additionally, we investigate the impact of basic differential privacy measures, such as adding Gaussian noise to the model parameters, to prevent the data exfiltration by compression attack. We also show how the attacker can make its attack resilient to differential privacy at the expense of decreasing the number of stolen images. Lastly, we propose an alternative prevention strategy by fine-tuning the model to be exported.

In this study, we evaluated and analyzed the effects of image compression on a neural network (NN)-based heart rate (HR) classification system. An NN-based HR-estimation system classifies facial images into groups of HR intervals. We evaluated the relationship between the image compression rates and accuracy of an NN-based HR estimation system. In our evaluation, the image of the face was compressed into lossless (PNG) and lossy (JPEG) formats to reduce the transmission bandwidth. The compressed images significantly reduce the required bandwidth and storage size. Furthermore, we analyzed the image classification accuracy of the DenseNet-121, VGG-16, and Inception V3 models. VGG-16 exhibited the highest performance, and the proposed system yielded an accuracy of 97.2% for correctly detecting the HR. Additionally, the results showed that lossy image compression quality slightly affected HR accuracy. This evaluation method can provide an effective solution under low computational complexity and low bitrate requirement for remote HR classification.

One of the major drawbacks of JPEG compression is that it cannot produce a good compression ratio for small mean square error (MSE). This paper presents a new method for lossy image compression for storing and transmitting images over the internet, which overcomes this weakness. The method comprises the following steps. The image is first converted in YUV color space and then partitioned in [Formula: see text] pixel blocks. The blocks are divided into four groups and principal component analysis (PCA) is applied to each group. The original pixel data is transformed in the new space, i.e. mode coefficients are calculated. The parameters of the PCA models and coefficients are converted to unsigned byte and saved in a binary file. Only 6 bits are used for storing the coefficients and eigenvectors’ coordinates range is adjusted according to the variance in each vector direction. Then Lempel–Ziv–Markov chain algorithm (LZMA) lossless compression is applied to obtain the final encoded file. The results show that the proposed method produced a much better compression ratio than JPEG for small MSEs. Two types of encoding schemes are evaluated and performance results are shown at the end of the paper.

This paper presents the interesting results of applying compression ratio (CR) in the prediction of the boundary between visually lossless and visually lossy compression, which is of particular importance in perceptual image compression. The prediction is carried out through the objective quality (peak signal-to-noise ratio, PSNR) and image representation in bits per pixel (bpp). In this analysis, the results of subjective tests from four publicly available databases are used as ground truth for comparison with the results obtained using the compression ratio as a predictor. Through a wide analysis of color and grayscale infrared JPEG and Better Portable Graphics (BPG) compressed images, the values of parameters that control these two types of compression and for which CR is calculated are proposed. It is shown that PSNR and bpp predictions can be significantly improved by using CR calculated using these proposed values, regardless of the type of compression and whether color or infrared images are used. In this paper, CR is used for the first time in predicting the boundary between visually lossless and visually lossy compression for images from the infrared part of the electromagnetic spectrum, as well as in the prediction of BPG compressed content. This paper indicates the great potential of CR so that in future research, it can be used in joint prediction based on several features or through the CR curve obtained for different values of the parameters controlling the compression.

Learned image lossy compression techniques have surpassed traditional methods in both subjective vision and quantitative evaluation. However, current models are only applicable to three-channel image formats, limiting their practical application due to the diversity and complexity of image formats. We propose a high-performance learned image compression model for general image formats. We first introduce a transfer method to unify any-channel image formats, enhancing the applicability of neural networks. This method's effectiveness is demonstrated through image information entropy and image homomorphism theory. Then, we introduce an adaptive attention residual block into the entropy model to give it better generalization ability. Meanwhile, we propose an evenly grouped cross-channel context module for progressive preview image decoding. Experimental results demonstrate that our method achieves state-of-the-art (SOTA) in the field of learned image compression in terms of PSNR and MS-SSIM. This work extends the applicability of learned image compression techniques to more practical production environments.

End-to-end optimization via deep neural networks has facilitated lossy image compression. Existing neural network-based entropy models for end-to-end optimized image compression are limited by parameterized Gaussian distributions with deterministic mean and variance and cannot achieve accurate rate estimation for bottleneck representation with varying statistics. In this paper, we propose a novel entropy model based on deep Gaussian process regression (DGPR) to address this problem. Specifically, the proposed entropy model leverages autoregressive DGPR to flexibly predict the channel-wise posterior distributions of high-dimensional bottleneck representation for entropy coding. Consequently, we develop a well-established bit-rate estimation scheme via posterior inference of DGPR using the learned probabilistic distribution. Furthermore, scalable training is achieved via tensor train decomposition and Monte Carlo sampling to enable tractable variational inference of DGPR. To our best knowledge, this paper is the first attempt to develop the learnable probabilistic model for flexible parameter estimation in entropy modeling. Experimental results show that the proposed model outperforms conventional image compression methods ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">e.g</i> ., JPEG2000 and BPG) as well as recent end-to-end optimized methods on the Kodak and Tecnick datasets in terms of rate-distortion performance.

Modern imaging systems produce a great volume of image data. In many practical situations, it is necessary to compress them for faster transferring or more efficient storage. Then, a compression has to be applied. If images are noisy, lossless compression is almost useless, and lossy compression is characterized by a specific noise filtering effect that depends on the image, noise, and coder properties. Here, we considered a modern HEIF coder applied to grayscale (component) images of different complexity corrupted by additive white Gaussian noise. It has recently been shown that an optimal operation point (OOP) might exist in this case. Note that the OOP is a value of quality factor where the compressed image quality (according to a used quality metric) is the closest to the corresponding noise-free image. The lossy compression of noisy images leads to both noise reduction and distortions introduced into the information component, thus, a compromise should be found between the compressed image quality and compression ratio attained. The OOP is one possible compromise, if it exists, for a given noisy image. However, it has also recently been demonstrated that the compressed image quality can be significantly improved if post-filtering is applied under the condition that the quality factor is slightly larger than the one corresponding to the OOP. Therefore, we considered the efficiency of post-filtering where a block-matching 3-dimensional (BM3D) filter was applied. It was shown that the positive effect of such post-filtering could reach a few dB in terms of the PSNR and PSNR-HVS-M metrics. The largest benefits took place for simple structure images and a high intensity of noise. It was also demonstrated that the filter parameters have to be adapted to the properties of residual noise that become more non-Gaussian if the compression ratio increases. Practical recommendations on the use of compression parameters and post-filtering are given.

Learned image compression has attracted considerable interests in recent years. An analysis transform and a synthesis transform, which can be regarded as coupled transforms, are used to encode an image to latent feature and decode the feature after quantization to reconstruct the image. Inspired by the success of invertible neural networks in generative modeling, invertible modules can be used to construct the coupled analysis and synthesis transforms. Considering the noise introduced in the feature quantization invalidates the invertible process, this paper proposes an Approximately Invertible Neural Network (A-INN) framework for learned image compression. It formulates the rate-distortion optimization in lossy image compression when using INN with quantization, which differentiates from using INN for generative modelling. Generally speaking, A-INN can be used as the theoretical foundation for any INN based lossy compression method. Based on this formulation, A-INN with a progressive denoising module (PDM) is developed to effectively reduce the quantization noise in the decoding. Moreover, a Cascaded Feature Recovery Module (CFRM) is designed to learn high-dimensional feature recovery from low-dimensional ones to further reduce the noise in feature channel compression. In addition, a Frequency-enhanced Decomposition and Synthesis Module (FDSM) is developed by explicitly enhancing the high-frequency components in an image to address the loss of high-frequency information inherent in neural network based image compression, thereby enhancing the reconstructed image quality. Extensive experiments demonstrate that the proposed A-INN framework achieves better or comparable compression efficiency than the conventional image compression approach and state-of-the-art learned image compression methods.

Machine learning, particularly deep learning, has revolutionized a number of fields, including medical diagnostics. In this study, federated learning is employed to address privacy concerns and data access limitations inherent in medical imaging. A simulated FL environment was used to investigate the performance of five pre-trained neural network models: DENSENET121, RESNET18, VGG-NET11, GOOGLENET and INCEPTION-V3. It emphasizes the optimization of training duration as well as the application of lossy image compression techniques such as JPEG in order to improve communication efficiency. We conducted a comparative analysis of the models' performance before and after image compression by evaluating the Area Under the Receiver Operating Characteristic Curve and the training time. According to the results, image compression can maintain or improve model performance while affecting training time, underscoring the trade-offs between model accuracy and computational efficiency.

Acquired remote sensing images can be noisy. This fact has to be taken into account in their lossy compression and classification. In particular, a specific noise filtering effect is usually observed due to lossy compression and this can be positive for classification. Classification can be also influenced by methodology of classifier learning. In this paper, we consider peculiarities of lossy compression of three-channel noisy images by better portable graphics (BPG) encoder and their further classification. It is demonstrated that improvement of data classification accuracy is not observed if a given image is compressed in the neighborhood of optimal operation point (OOP) and the classifier training is performed for the noisy image. Performance of neural network based classifier is studied. As demonstrated, its training for compressed remote sensing data is able to provide certain benefits compared to training for noisy (uncompressed) data. Examples for Sentinel data used in simulations are offered.

Introduction. Lossy image compression algorithms play a crucial role in various domains, including graphics, and image processing. As image information density increases, so do the resources required for processing and transmission. One of the most prominent approaches to address this challenge is color quantization, proposed by Orchard et al. (1991). This technique optimally maps each pixel of an image to a color from a limited palette, maintaining image resolution while significantly reducing information content. Color quantization can be interpreted as a clustering problem (Krishna et al. (1997), Wan (2019)), where image pixels are represented in a three-dimensional space, with each axis corresponding to the intensity of an RGB channel. The purpose of the paper. Scaling of traditional algorithms like K-Means can be challenging for large data, such as modern images with millions of colors. This paper reframes color quantization as a three-dimensional stochastic transportation problem between the set of image pixels and an optimal color palette, where the number of colors is a predefined hyperparameter. We employ Stochastic Quantization (SQ) with a seeding technique proposed by Arthur et al. (2007) to enhance the scalability of color quantization. This method introduces a probabilistic element to the quantization process, potentially improving efficiency and adaptability to diverse image characteristics. Results. To demonstrate the efficiency of our approach, we present experimental results using images from the ImageNet dataset. These experiments illustrate the performance of our Stochastic Quantization method in terms of compression quality, computational efficiency, and scalability compared to traditional color quantization techniques. Conclusions. This study introduces a scalable algorithm for solving the color quantization problem without memory constraints, demonstrating its efficiency on a subset of images from the ImageNet dataset. The convergence speed of the algorithm can be further enhanced by modifying the update rule with alternative methods to Stochastic Gradient Descent (SGD) that incorporate adaptive learning rates. Moreover, the stochastic nature of the proposed solution enables the utilization of parallelization techniques to simultaneously update the positions of multiple quants, potentially leading to significant performance improvements. This aspect of parallelization and its impact on algorithm efficiency presents a topic for future research. The proposed method not only addresses the limitations of existing color quantization techniques but also opens up new possibilities for optimizing image compression algorithms in resource-constrained environments. Keywords: non-convex optimization, stochastic optimization, stochastic quantization, color quantization, lossy compression.

The object of the study is the process of lossy compression of noisy images and their post-filtering. The subject of the study is the approach to efficient two-stage processing (compression and post-filtering) for better portable graphics (BPG) coder and prediction of its efficiency. The goal of the study is to analyze performance characteristics of the considered two-stage approach and to propose an approach to their prediction. Methods used: numerical simulation, regression, statistical analysis. Results obtained: 1) the considered approach advantage is that it is able to provide improvement of quality of compressed noisy image under condition that an image is compressed with compression ratio smaller than that one corresponding to optimal operation point; 2) the approach efficiency depends on several factors including noise intensity, image complexity, and filter type and parameters; 3) the main characteristics of the two-step procedure can be quite accurately predicted in advance and this allows offering useful information for decision undertaking on what value of the coder parameter to apply; 4) this leads to either improving the compressed and processed image quality compared to its original version or, at least, to avoiding quality degradation. Conclusions: based on the results of the study, it is worth 1) predicting performance characteristics for the two-stage processing; 2) adapting the processing to image complexity and noise intensity.

In this paper, we introduce an efficient lossy coding procedure specifically tailored for handling video sequences of automotive high-dynamic range (HDR) image sensors in advanced driver-assistance systems (ADASs) for autonomous vehicles. Nowadays, mainly for security reasons, lossless compression is used in the automotive industry. However, it offers very low compression rates. To obtain higher compression rates, we suggest using lossy codecs, especially when testing image processing algorithms in software in-the-loop (SiL) or hardware-in-the-loop (HiL) conditions. Our approach leverages the high-quality VP9 codec, operating in two distinct modes: grayscale image compression for automatic image analysis and color (in RGB format) image compression for manual analysis. In both modes, images are acquired from the automotive-specific RCCC (red, clear, clear, clear) image sensor. The codec is designed to achieve a controlled image quality and state-of-the-art compression ratios while maintaining real-time feasibility. In automotive applications, the inherent data loss poses challenges associated with lossy codecs, particularly in rapidly changing scenes with intricate details. To address this, we propose configuring the lossy codecs in variable bitrate (VBR) mode with a constrained quality (CQ) parameter. By adjusting the quantization parameter, users can tailor the codec behavior to their specific application requirements. In this context, a detailed analysis of the quality of lossy compressed images in terms of the structural similarity index metric (SSIM) and the peak signal-to-noise ratio (PSNR) metrics is presented. With this analysis, we extracted some codec parameters, which have an important impact on preservation of video quality and compression ratio. The proposed compression settings are very efficient: the compression ratios vary from 51 to 7765 for grayscale image mode and from 4.51 to 602.6 for RGB image mode, depending on the specified output image quality settings. We reached 129 frames per second (fps) for compression and 315 fps for decompression in grayscale mode and 102 fps for compression and 121 fps for decompression in the RGB mode. These make it possible to achieve a much higher compression ratio compared to lossless compression while maintaining control over image quality.

FPWT: Filter pruning via wavelet transform for CNNs

Recent advancements in 3D data capture have enabled the real-time acquisition of high-resolution 3D range data, even in mobile devices. However, this type of high bit-depth data remains difficult to efficiently transmit over a standard broadband connection. The most successful techniques for tackling this data problem thus far have been image-based depth encoding schemes that leverage modern image and video codecs. To our knowledge, no published work has directly optimized the end-to-end losses of a depth encoding scheme sandwiched around a lossy image compression codec. We present N-DEPTH, a compression-resilient neural depth encoding method that leverages deep learning to efficiently encode depth maps into 24-bit RGB representations that minimize end-to-end depth reconstruction errors when compressed with JPEG. N-DEPTH’s learned robustness to lossy compression expands to video codecs as well. Compared to an existing state-of-the-art encoding method, N-DEPTH achieves smaller file sizes and lower errors across a large range of compression qualities, in both image (JPEG) and video (H.264) formats. For example, reconstructions from N-DEPTH encodings stored with JPEG had dramatically lower error while still offering 29.8%-smaller file sizes. When H.264 video was used to target a 10 Mbps bit rate, N-DEPTH reconstructions had 85.1%-lower root mean square error (RMSE) and 15.3%-lower mean absolute error (MAE). Overall, our method offers an efficient and robust solution for emerging 3D streaming and 3D telepresence applications, enabling high-quality 3D depth data storage and transmission.

Lossy compression of remote-sensing images is a typical stage in their processing chain. In design or selection of methods for lossy compression, it is commonly assumed that images are noise-free. Meanwhile, there are many practical situations where an image or a set of its components are noisy. This fact needs to be taken into account since noise presence leads to specific effects in lossy compressed data. The main effect is the possible existence of the optimal operation point (OOP) shown for JPEG, JPEG2000, some coders based on the discrete cosine transform (DCT), and the better portable graphics (BPG) encoder. However, the performance of such modern coders as AVIF and HEIF with application to noisy images has not been studied yet. In this paper, analysis is carried out for the case of additive white Gaussian noise. We demonstrate that OOP can exist for AVIF and HEIF and the performance characteristics in it are quite similar to those for the BPG encoder. OOP exists with a higher probability for images of simpler structure and/or high-intensity noise, and this takes place according to different metrics including visual quality ones. The problems of providing lossy compression by AVIF or HEIF are shown and an initial solution is proposed. Examples for test and real-life remote-sensing images are presented.

Lossy image compression is a fundamental technology in media transmission and storage. Variable-rate approaches have recently gained much attention to avoid the usage of a set of different models for compressing images at different rates. During the media sharing, multiple re-encodings with different rates would be inevitably executed. However, existing Variational Autoencoder (VAE)-based approaches would be readily corrupted in such circumstances, resulting in the occurrence of strong artifacts and the destruction of image fidelity. Based on the theoretical findings of preserving image fidelity via invertible transformation, we aim to tackle the issue of high-fidelity fine variable-rate image compression and thus propose the Invertible Continuous Codec (I2C). We implement the I2C in a mathematical invertible manner with the core Invertible Activation Transformation (IAT) module. I2C is constructed upon a single-rate Invertible Neural Network (INN) based model and the quality level (QLevel) would be fed into the IAT to generate scaling and bias tensors. Extensive experiments demonstrate that the proposed I2C method outperforms state-of-the-art variable-rate image compression methods by a large margin, especially after multiple continuous re-encodings with different rates, while having the ability to obtain a very fine variable-rate control without any performance compromise.

Classical vs. neural network-based PCA approaches for lossy image compression: Similarities and differences

The compression of images for efficient storage and transmission is crucial in handling large data volumes. Lossy image compression reduces storage needs but introduces perceptible distortions affected by content, compression levels, and display environments. Each compression method generates specific visual anomalies like blocking, blurring, or color shifts. Standardizing efficient lossy compression necessitates evaluating perceptual quality. Objective measurements offer speed and cost efficiency, while subjective assessments, despite their cost and time implications, remain the gold standard. This paper delves into essential research queries to achieve visually lossless images. The paper describes the influence of compression on image quality, appropriate objective image quality metrics (IQMs), and the effectiveness of subjective assessment methods. It also provides an overview of the existing literature, surveys, and subjective and objective image quality assessment (IQA) methods. Our aim is to offer insights, identify challenges in existing methodologies, and assist researchers in selecting the most effective assessment approach for their needs.