Search

The rise of online sharing platforms has provided people with diverse and convenient ways to share images. However, a substantial amount of sensitive user information is contained within these images, which can be easily captured by malicious neural networks. To ensure the secure utilization of authorized protected data, reversible adversarial attack techniques have emerged. Existing algorithms for generating adversarial examples do not strike a good balance between visibility and attack capability. Additionally, the network oscillations generated during the training process affect the quality of the final examples. To address these shortcomings, we propose a novel reversible adversarial network based on generative adversarial networks (RA-RevGAN). In this paper, the generator is used for noise generation to map features into perturbations of the image, while the region selection module confines these perturbations to specific areas that significantly affect classification. Furthermore, a robust attack mechanism is integrated into the discriminator to stabilize the network’s training by optimizing convergence speed and minimizing time cost. Extensive experiments have demonstrated that the proposed method ensures a high image generation rate, excellent attack capability, and superior visual quality while maintaining high classification accuracy in image restoration. [ABSTRACT FROM AUTHOR]

This study aims to evaluate the therapeutic potential of cationic liposomal neostigmine bromide (NB), a novel drug delivery system, for the treatment of detrusor underactivity. By comparing the characteristics of NB‐liposomes (NLP), NB‐β‐cyclodextrin inclusion complex liposomes (NCLP), and NB‐mesoporous silica nanoparticle@CaCO3 liposomes (NMCLP), NMCLP is selected as the main research subject. It has an average particle size and zeta potential of 100 nm and +50 mV, and its encapsulation efficiency and loading capacity of NB are 14.75% and 12.8%, respectively. Most importantly, NMCLP shows the best in vitro release performance among the three liposomes, demonstrating its ability in sustained release of NB. During cell and animal assays, efficient cellular uptake of liposomes through liposome‐specific pathways is observed, facilitating targeted drug delivery, and in vivo experiments demonstrate the efficacy of NMCLP in improving bladder function in mice. Urodynamic measurements show increased bladder capacity and reduced voiding pressure, indicating enhanced bladder muscle activity. Histological analysis reveals the distribution and deep penetration of NMCLP within bladder tissues, supporting its localized drug effect. Therefore, NMCLP holds promise as a targeted and effective therapeutic strategy for detrusor underactivity. [ABSTRACT FROM AUTHOR]

In recent years, there has been an increasing concern regarding the security and copyright issues associated with images. Unscrupulous individuals exploit electronic images through theft and manipulation, subsequently disseminating them widely, thereby posing significant threats to image copyright ownership and reliability. To address potential screen-shooting attacks on images, this paper proposes a robust watermarking algorithm based on entropy-weighted Harris corner detection and adaptive embedding radius. Firstly, this algorithm uses entropy as the weighted coefficient of Harris corner response values to extract feature points with rich texture features and high robustness. Then, the SIFT algorithm is used to assign orientations to feature points and construct and filter out non-overlapping feature regions. Next, the embedding radius of the watermark is adaptively selected based on the average gray value of the image feature region, and the watermark is embedded into the DFT coefficients of the image feature region. Finally, after completing the watermark embedding in all feature regions, the watermarked image is obtained. In the watermark extraction stage, the captured image is first geometrically distorted and then processed with a Gaussian function before extracting the watermark information from the image. Experimental results show that this algorithm has strong robustness against screen-shooting attacks and common image attacks, while maintaining high transparency. [ABSTRACT FROM AUTHOR]

Benefiting from its high thermal conductivity, simple structure, and light weight, the pulsating heat pipe (PHP) can meet the requirements for high efficiency, flexibility, and low cost in industrial heat transfer applications such as aerospace detector cooling and vehicle thermal management. Compared to a PHP working at room temperature, the mechanism of a PHP with hydrogen as the working fluid differs significantly due to the unique thermal properties of hydrogen. In this paper, a two-dimensional model of a hydrogen PHP with a filling ratio of 51% was established to study the flow characteristics and thermal performance. The volume of fluid (VOF) method was used to capture the phase distribution and interface dynamics, and the Lee model was employed to account for phase change. To validate the model, a comparison was conducted between the simulation results and experimental data obtained in our laboratory. The simulation results show that the pressure and temperature errors were within 25% and 5%, respectively. Throughout a pressure oscillation cycle, the occurrence of uniform flow velocity, acceleration, and flow reversal can be attributed to the changes in the vapor–liquid phase distribution resulting from the effect of condensation and evaporation. In addition, when the fluid velocity was greater than 0.6 m/s, dynamic contact angle hysteresis was observed in the condenser. The results contribute to a deeper understanding of the flow and heat transfer mechanism of the hydrogen PHPs, which have not been yet achieved through visualization experiments. [ABSTRACT FROM AUTHOR]

Detection of double Joint photographic experts group (JPEG) compression is a major part of image forensics, especially double JPEG compression detection with the same quantization matrix is still a challenging problem. Detection of double JPEG compression for different quantization matrices has achieved good performance so far, and some methods for detecting the same quantization matrix have been proposed. However, most methods designed for grayscale images, where they utilize only truncation and rounding errors, not fully exploiting all the useful information of color JPEG images. In addition, there is no suitable model to analyze the convergence of JPEG images. These results in poor accuracy of the double compression detection for color JPEG images with the same quantization matrix. To solve this problem, we propose an effective method that utilizes third-order polynomial fitting for error convergence curves. Firstly, five successive compressions are used to extract the quantization error, dequantization error, coefficient error, and truncation error in each compression and decompression process. Secondly, based on the convergence characteristics of these four types of error information, a third-order polynomial model and least squares are applied for the fit, and then, the slope of each point of the fitted curve and the difference of the slopes are used as convergence characteristics. Moreover, the fitted curves predict the pre-compression state, which can effectively amplify the gap between single and double compression. Finally, support vector machines (SVM) are used to train and predict the final features. The experiments illustrate that the proposed method enhances accuracy by 2–3% compared to the most recent method and by approximately 15% compared to the earlier method in multiple datasets, including NRCS, MIXDATA, and MIXDATA-I. [ABSTRACT FROM AUTHOR]

Advances in mobile communication have enabled images to be shared more conveniently, and the consequent risk of privacy breaches has been increasingly emphasized. Previously, traditional image encryption algorithms were mostly adopted by individuals to encrypt plain images into noise-like meaningless cipher images. This type of algorithm provides security for image privacy protection but is unable to achieve usability in cloud platforms. Furthermore, existing image encryption algorithms only target the protection of lossless compression or non-compressed formats images (e.g., BMP, PNG, etc.), but neglect the encryption needs of JPEG format images. Since the latter has the benefits of less file storage and excellent visual quality, it has been widely applied in a variety of scenarios. Thus, a thumbnail preserving encryption (TPE) scheme based on Markov model (TPE-MM) for JPEG images is proposed to equilibrate security and availability. Specifically, an approximate sum-preserving encryption method (ASP-EM) in restricted range based on its features is proposed to preserve the plain thumbnail features to accomplish cipher image usability. Subsequently, an adaptive encryption method based on Markov model (AEM-MM) was proposed to improve the encryption effect of cipher images and retain the compression performance of JPEG images. Simulation experiments show that the cipher image has good visual quality, low expansion rate, and is resistant to outline attacks. [ABSTRACT FROM AUTHOR]

The traditional image encryption method is to encrypt the plain image (PLI) into a noise-like encrypted image (EI), which can hide the information of the PLI well. However, when the EI is transmitted over the Internet, it is easy to attract the attention of hackers due to its special appearance, and the probability of EI being attacked is also greatly improved. To solve this problem, a visually secure image encryption scheme (VSIES) is proposed in this paper by using the newly designed one-dimensional sinusoidal chaotic (1-DSC) map, P-tensor product compressive sensing and discrete U transform (DUT). First, a key generation mechanism is designed where the secure hash algorithm SHA-512 is used to generate the key parameters. Then, this key is used to control the 1-DSC map to generate a measurement matrix, and the sparse coefficient matrix of the PLI is measured. Next, a double zigzag confusion method is designed to scramble the measurement values and the obtained image is diffused to find the secret image (SE). Finally, in order to randomly embed SE into a carrier image to obtain a cipher image with visually meaningful, a DUT-based embedding technique is proposed. The experimental results show that the information entropy value of the SE generated by the encryption scheme is close to 8, which proves that our encryption scheme has a good encryption effect. The visual security (VIS) of our encrypted images is 10–15 dB higher than some existing VSIES, which fully proves that our proposed encryption scheme has high VIS. [ABSTRACT FROM AUTHOR]

Chaotic systems have played a vital role within the domain of chaos-based image encryption, owing to its significant advantages including sensitivity to initial values, ergodicity and others. Nonetheless, the current low-dimensional and less-scroll chaotic systems exhibit low complexity and randomness, leading to unsatisfactory security performance of the associated image encryption schemes. To settle the aforementioned issue, this paper introduces a novel hyperchaotic system with five scrolls and explores its application in visually meaningful image encryption (VMIE). Firstly, a new 5-scroll Rossler hyperchaotic system (5-SRHS) is designed building upon the principle of translation transformation (PTT), using the Rossler chaotic system (RCS) as a foundation. The performance of the 5-SRHS is extensively analyzed using various techniques, including but not limited to phase diagrams, Lyapunov exponents, and the NIST test. Subsequently, based on the 5-SRHS, a VMIE method is introduced. Specifically, to improve image reconstruction quality and reasonable allocation of sampling resources, a sub-region adaptive block compressive sensing (SR-ABCS) driven by image texture features is presented and applied to pre-encrypt the plain image, resulting in the generation of the encrypted image. Next, the obtained encrypted image is dynamically embedded into the host image utilizing improved grayscale difference statistics (IGDS) and 2K correction, enhancing imperceptibility of the steganographic image. Moreover, SHA-256 is integrated with the 5-SRHS to generate plaintext-related random sequences, further achieving higher security. Simulation experiments validate that the proposed scheme surpasses existing state-of-the-art (SOTA) schemes regarding steganography and reconstruction quality. Moreover, it is proven not only for secure transmission of natural images but also for medical images. [ABSTRACT FROM AUTHOR]

Photothermal immunotherapy can provide an efficient and convenient strategy for cancer treatment by inducing immunogenic cell death (ICD) to generate vaccine‐like functions in situ, which highly relies on suitable photothermal agents with superior photothermal effect. Herein, a novel diradical‐featured organic small‐molecule photothermal material (TTD‐TPE) based on 4,6‐di(2‐thienyl)thieno[3,4‐c][1,2,5]thiadiazole and tetraphenylethylene is developed to exhibit near‐infrared (NIR) absorption of 808 nm, high photothermal conversion efficiency (PTCE) of about 75%, and superior photoacoustic imaging (PAI) property in nanoparticle. The excellent photothermal effect of TTD‐TPE nanoparticles originates from the integrated synergic effect of D–A interactions, diradical character, and excited‐state intramolecular motion, which not only ablates the primary tumor by localized high temperature but also transforms the primary tumor into vaccine in situ through ICD to inhibit the growth of distant and metastatic tumors, demonstrating the huge potential in photothermal immunotherapy. Thus, this finding encourage researchers to develop superior organic small‐molecule photothermal materials to boost the simple but useful modality for cancer treatment. [ABSTRACT FROM AUTHOR]

Magnetic resonance (MR) angiography is one of the main diagnostic approaches for cardiac-cerebral vascular diseases. Nevertheless, the non-contrast-enhanced MR angiography suffers from its intrinsic problems derived from the blood flow-dependency, while the clinical Gd-chelating contrast agents are limited by their rapid vascular extravasation. Herein, we report a hypersensitive MR angiography strategy based on interlocking stratagem of zwitterionic Gd-chelate contrast agents (PAA-Gd). The longitudinal molar relaxivity of PAA-Gd was 4.6-times higher than that of individual Gd-chelates as well as appropriate blood half-life (73.8 min) and low immunogenicity, enabling sophisticated micro-vessels angiography with a resolution at the order of hundred micrometers. A series of animal models of cardiac-cerebrovascular diseases have been built for imaging studies on a 7.0 T MRI scanner, while the clinical translation potential of PAA-Gd has been evaluated on swine on a 3.0 T clinical MRI scanner. The current studies offer a promising strategy for precise diagnosis of vascular diseases. Current contrast-enhanced magnetic resonance angiography approaches are sub-optimal. Here the authors present a hypersensitive MR angiography strategy based on interlocking stratagem of zwitterionic Gd-chelate contrast agents (PAA-Gd), enabling sophisticated micro-vessel angiography of cardiac-cerebrovascular diseases with ultrahigh resolution. [ABSTRACT FROM AUTHOR]

Central nervous system (CNS) diseases have been widely acknowledged as one of the major healthy concerns globally, which lead to serious impacts on human health. There will be about 135 million CNS diseases cases worldwide by mid-century, and CNS diseases will become the second leading cause of death after the cardiovascular disease by 2040. Most CNS diseases lack of effective diagnostic and therapeutic strategies with one of the reasons that the biological barrier extremely hampers the delivery of theranostic agents. In recent years, nanotechnology-based drug delivery is a quite promising way for CNS diseases due to excellent properties. Among them, cell membrane-based nanomaterials with natural bio-surface, high biocompatibility and biosafety, are of great significance in both the diagnosis and treatment of different CNS diseases. In this review, the state of art of the fabrication of cell membranes-based nanomaterials is introduced. The characteristics of different CNS diseases, and the application of cell membranes-based nanomaterials in the theranostics are summarized. In addition, the future prospects and limitations of cell membrane nanotechnology are anticipated. Through summarizing the state of art of the fabrication, giving examples of CNS diseases, and highlighting the applications in theranostics, the current review provides designing methods and ideas for subsequent cell membrane nanomaterials. [ABSTRACT FROM AUTHOR]

A series of four‐armed poly(ϵ‐caprolactone‐co‐d,ʟ‐lactide)‐block‐poly(d‐lactide) (4a‐CDL‐D) copolymers with different poly(d‐lactide) (PDLA) block lengths (Mn,PDLA) were synthesized via ring opening polymerization using pentaerythritol as initiator. They were blended with poly(ʟ‐lactide) (PLLA) to improve its crystallization rate and mechanical properties. It was found that the crystallization of PLLA in the PLLA/4a‐CDL‐D blend was affected differently when Mn,PDLA changed. In the PLLA/4a‐CDL blend without PDLA blocks, the crystallization was retarded obviously because of the dilution effect of 4a‐CDL on PLLA crystallization. Interestingly, a crystallization acceleration effect was found when Mn,PDLA was 0.5k although no stereocomplex (SC) crystallites existed. When Mn,PDLA was 1k or higher (1.5k and 2k), SC crystallites could be formed in the PLLA/4a‐CDL‐D blends. However, when Mn,PDLA was 1k or 1.5 k, the formation of SC crystallites was difficult and they showed weaker acceleration effect on PLLA crystallization. While when Mn,PDLA was 2k, the crystallization of PLLA was accelerated markedly due to the significant increase of nucleation sites. At last, these blends had longer elongation at breaks than neat‐PLLA. This work provided a novel method for the acceleration of PLLA crystallization by adjusting the PDLA chain length in PLLA blends. [ABSTRACT FROM AUTHOR]

Copyright of Frontiers of Information Technology & Electronic Engineering is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The cost of collecting and labeling open-sourced datasets which promote the development of deep learning is expensive. Thus, it is important to design an efficient open-sourced data set protection algorithm to detect the open-sourced data sets used in illegal ways. In this paper, a protection algorithm based on adaptive robust blind-watermark is proposed suitable for multiple paired open-sourced datasets, and the evaluation criteria of the algorithm are defined. Specifically, in the embed stage, the highly concealed of the watermark is realized by combining UNet and double GAN to take into account the local and global features of the carrier and the watermark image. A preprocess network is used in the Embedding network to adapt different watermark size. In the extraction part, a modified feature sharing UNet with GAN is used to ensure robustness of the extraction network. Paired datasets are used for training to ensure accurate extraction of watermarks. After the target model is trained using the watermarked dataset, its inference output will contain watermark information. When it is believed that a suspicious model is illegally trained with the dataset, it can be verified by the watermark extracted from inference output of the suspicious model. We evaluate our method on three target models including nine datasets. The results show that our framework successfully verifies the dataset used illegally and without a noticeable impact on the target model task when training with the watermark dataset. [ABSTRACT FROM AUTHOR]

Malignant tumors have been one of the major reasons for deaths worldwide. Timely and accurate diagnosis as well as effective intervention of tumors play an essential role in the survival of patients. Genomic instability is the important foundation and feature of cancer, hence, in vivo oncogene imaging based on novel probes provides a valuable tool for the diagnosis of cancer at early-stage. However, the in vivo oncogene imaging is confronted with great challenge, due to the extremely low copies of oncogene in tumor cells. By combining with various novel activatable probes, the molecular imaging technologies provide a feasible approach to visualize oncogene in situ, and realize accurate treatment of tumor. This review aims to declare the design of nanoprobes responded to tumor associated DNA or RNA, and summarize their applications in detection and bioimaging for tumors. The significant challenges and prospective of oncogene-targeting nanoprobes towards tumors diagnosis are revealed as well. [ABSTRACT FROM AUTHOR]

Deep neural networks (DNNs) are prone to produce incorrect prediction results under the attack of adversarial samples. To cope with this problem, some defense methods are presented. However, most of them are based on adversarial training, which has great computational consumption and does not start from strengthening the architecture of the network model itself to resist the adversarial attack. Recent studies have shown that feature denoising can remove the adversarial perturbations in the adversarial samples. In this paper, we propose a lightweight denoising network with residual connection (LDN-RC), on which the internal denoising block and the intermediate denoising block are introduced for feature denoising and sample denoising, respectively; the two denoising blocks are combined in the network model, which can withstand the interference of the adversarial perturbations in the adversarial samples to a large extent and also save computational resources. In the training strategy, a two-stage denoising approach and fine-tuning are presented to train the RESNET network model on MNIST, CIFAR-10, and SVHN datasets, and the accuracy of the enhanced network model exceeds 60% on all three datasets under the -PGD white-box attack, which demonstrate that LDN-RC can effectively improve the adversarial robustness of the network model. [ABSTRACT FROM AUTHOR]

With the development of cloud storage, more and more users upload images to the cloud. However, images stored in the cloud face the risk of unauthorized data mining by cloud service providers and being stolen by hackers. Encryption can protect image privacy, but traditional image encryption algorithms sacrifice image usability for security. To protect privacy while preserving image usability, two approximate thumbnail-preserving encryption (TPE) schemes, called dynamic range preserving encryption (DRPE) and approximate TPE with LSB Embedding (TPE-LSB), have been presented by Marohn in 2017. However, there is the possibility of decryption failure for DRPE, the cipher image robustness is poor for TPE-LSB, which cannot resist noise attacks. Additionally, both methods have the problems of the high file expansion rate of cipher image and poor perceptual quality of cipher image thumbnail. To solve these problems, a multi-level DWT information self-embedding method for thumbnail preserving encryption (TPE-ISE) is proposed. Compared with the previous works, the TPE-ISE scheme achieves a controllable compression ratio of cipher images, the perceptual quality of cipher images is closer to that of plain images, and the ability of cipher images to resist noise attacks is stronger. A series of experiments verify the superiority of the proposed algorithm. [ABSTRACT FROM AUTHOR]

Encryption of meaningful images into unidentifiable noise-like images can improve the security of images during storage and transmission. In this paper, a novel color image encryption method based on the lifting scheme and cross-component permutation (CIE-LSCP) is introduced. First, original image is divided into red, green and blue components, and then the three components are processed by a preprocessing strategy based on the lifting scheme (PSLS) to change the statistical distribution of the pixels. Second, a block-based cross-component permutation based on index vectors (BCPIV) is presented to divide three preprocessed components into blocks and perform the cross-component permutation operation on them, and three random matrices are utilized to determine the target component, target block and target pixel position of the current pixel to be moved, respectively. Subsequently, a multi stochastic diffusion based on random sequences (MSDRS) is developed to alter the pixel values of the permutated components, and finally a ciphertext image is gotten by merging the three components. Moreover, the SHA256 hash values of the plaintext image are used to obtain the initial parameters of the chaotic system, and the obtained chaotic sequences are applied in the image encryption process. Wherein the generated random sequences are highly reliant on the plaintext image, making the encryption scheme resistant to both known-plaintext attacks and chosen-plaintext attacks. Experimental results demonstrate that the proposed scheme has good security and effectiveness and can be applied for secure transmission of digital images over the Internet. [ABSTRACT FROM AUTHOR]

Recently, most histogram shifting-based reversible data hiding (RDH) algorithms have considered the impact of textural information on embedding performance, while exploiting pixels with different local complexities. Prioritizing pixels with small local complexities to accommodate secret data decreases the invalid shifting pixels, thereby reducing distortion. However, though effective, the local complexity is not calculated precisely enough, which results in inaccurate texture division and does not considerably reduce distortion. Thus, we employ a novel local complexity calculation and multipeak embedding (MPE) to effectively improve the capacity-distortion performance. Specifically, the host image is first preprocessed by the dot-cross pattern and divided into two subsets. Then the pixel local complexity of each subset is computed by using the spatial correlation of pixels (SCOP) to improve calculation accuracy. Finally, the peak bins to be expanded in the regions with lower local complexity are adaptively selected for embedding with secret data by MPE. To ensure that the authorized operator can securely gain error-free secret data, we design the location index sequences as special keys, which guarantee the algorithm reversibility and enhance the security of the algorithm at the same time. Experimental results show that our algorithm has superior embedding effectiveness compared to some state-of-the-art RDH methods. [ABSTRACT FROM AUTHOR]

The hypoxic nature of solid tumors limits the efficacy of radiotherapy (RT) and leads to radiation resistance. Hypoxic radiosensitizers can enhance tumor radiosensitivity by mimicking the effects of oxygen, but their efficacy has been limited by the heterogeneous oxygen distribution in tumor tissue. Herein, a multimodal therapeutic nanoplatform fabricated by co‐encapsulation of chlorin e6 and tirapazamine (TPZ) with an amphiphilic polymeric conjugate of cisplatin (CDDP) and metronidazole is reported. This platform could kill the tumor periphery cells by the deeply penetrated oxygen‐consuming sonodynamic therapy and unify the heterogeneously hypoxic context simultaneously, which then actuate the release and activation of the loaded TPZ. TPZ could further sensitize RT along with CDDP and metronidazole residues under the resultant hypoxic condition, which significantly decreases the radiation dosage required to cause massive cell damage. Except for the significantly suppressed tumor growth and metastasis caused by the multimodal therapeutic nanoplatform, its hypoxia‐directed feature also endows it with excellent safety. [ABSTRACT FROM AUTHOR]

With the advancement of multimedia technology and coming of big data era, the size of image data is significantly increased. However, the traditional image encryption methods cannot solve the emerging problems of efficient compression. To settle with this challenge, an effective content-adaptive image compression and encryption method based on compressive sensing and double random phase encoding (DRPE) is proposed in this paper. The original image is converted to one low-frequency part and three high-frequency parts by DWT and then permutated by sorting-based chaotic sequences. Afterward, a novel measurement matrix optimization algorithm based on adaptive step size is presented to measure the high-frequency components. To enhance the security of the scheme, the DRPE, quantization, and diffusion are successively performed on the complex matrix composed of the shuffled low-frequency component and three measurement value matrices to obtain the cipher image. Logistic-Sine chaotic system is utilized to produce the chaotic keystreams for the encryption process, and its system parameter and initial value are determined by the information entropy of the plain image and external key parameters, so that the proposed cipher can withstand known-plaintext and chosen-plaintext attacks effectively. Numerical experiments demonstrate the effectiveness of the proposed image compression and encryption algorithm. [ABSTRACT FROM AUTHOR]

In this paper, we propose a new image compression and encryption scheme (ICES) based on compressed sensing (CS) and double random phase encoding (DRPE), as well as a joint decryption on the cloud and user side. In the encryption, the plain image is firstly decomposed into approximate component and detail components by discrete wavelet transform (DWT), then the approximate component is scrambled, and detail components are compressed using a measurement matrix constructed by the Logistic-tent system. Secondly, the scrambled approximate component and compressed detail components are combined into a complex matrix, and then it is subjected to DRPE to get the amplitude and phase matrices. Subsequently, the resulting matrices are performed random pixel scrambling and diffusion to obtain the final cipher image. During the decryption, the cipher image is first partially decrypted on the cloud, and then fully decrypted to recover the original image on the user side, and we can judge whether the cloud has cheated us by comparing the contour similarity between the approximate component and the detail component returned by the cloud to the user, which not only significantly shortens the decryption time, but also effectively prevents malicious deception of the cloud. Experimental results and performance analyses demonstrate its effectiveness and efficiency. [ABSTRACT FROM AUTHOR]

Visually meaningful image encryption may keep the data security and appearance security of the digital images. However, there are still security and efficiency shortcomings existing in the current algorithms. To solve these problems, we propose an effective visually meaningful color image encryption scheme by combining hybrid multi-objective particle swarm optimization (HMPSO), block compressed sensing (BCS) and Hessenberg decomposition (HD). Firstly, the R, G, B components of color image are segmented averagely and represented sparsely by discrete cosine transform (DCT), respectively. Next, the obtained sparse images are scrambled by the use of zigzag path and measured by BCS to obtain the measurement value matrices. To improve its security, the key associated with the plain image is used as the initial value of the nonlinear chaotic system Henon, and a cross-component dislocation and diffusion strategy are applied to the measurements using the chaotic sequences generated by Henon to obtain the secret image, which enhances the ability of the algorithm to resist chosen-plaintext attack. Subsequently, the secret images are fused into the carrier image by the HD embedding algorithm to generate the final visually meaningful cipher image. In addition, in order to enhance the quality of the reconstructed image and cipher image, HMPSO is implemented to optimize the threshold value of sparse coefficient modification and the embedding rate simultaneously. Simulation results and performance analysis demonstrate the effectiveness, confidentiality and robustness of the proposed scheme. [ABSTRACT FROM AUTHOR]

In this paper, a robust compressed sensing image encryption algorithm based on generative adversarial network, convolutional neural network (CNN) denoising network and chaotic system is developed. Firstly, we use a sampling network to get the measurement of plain image. Second, the cipher image is obtained by scrambling the measurement through the Logistic-Tent chaotic system. After getting cipher image, the decryption party obtains decrypted measurement by inverse scrambling the cipher image, and then sends it to the reconstruction network to obtain decrypted reconstructed image. Finally, by using the CNN denoiser, the image quality and the visual expression of final decrypted image can be improved. In this scheme, the dual denoiser structure based on reconstruction network and CNN denoiser can effectively resist noise attacks. Besides, the proposed training strategy with noise injection can further improve the robustness of network. Experiments show our method has high reconstruction quality, efficiency, robustness and security. [ABSTRACT FROM AUTHOR]

In this paper, an image cipher is presented based on DNA sequence operations, image filtering and memrisitve chaotic system. Firstly, plain image is preprocessed by a self-updating transformation based on dynamic image filtering (STDIF), which may solve the problem that conventional filtering is invalid for special image such as all-zeros. Subsequently, the resulting image is converted into the DNA sequences by the dynamic DNA encoding rules, and the rules are produced by plain image information and chaotic system. Next, the DNA matrix is permutated by a double random 3D matrix scrambling (DR3DMS), thus every element may arbitrarily move to other position. Thereafter, the permutated matrix is diffused by a plane diffusion of 3D DNA matrix controlled by plain image information (PD3DPI) to resist statistical attacks. Finally, the obtained DNA matrix is transformed into cipher image via DNA decoding rules. Additionally, SHA 256 hash function value of plain image is applied to choose keystreams from the obtained chaotic sequences, which not only solves the problem of repeatedly generating chaotic sequences when encrypting different images, but also results into higher difficulty for hackers to decipher the method by use of plaintext attack. Experimental results and security analyses demonstrate its security and effectiveness. [ABSTRACT FROM AUTHOR]

Compared to 1D compressed sensing (CS), 2D CS is more efficient for compressing the plaintext image from two directions, but security level of current 2D CS-based ciphers is unsatisfactory. To solve this problem, this paper presents a novel color image compression and encryption algorithm by combining 2D CS, information entropy and chaos. Firstly, the color image is decomposed into red, green and blue components, then they are sparsely transformed by the discrete wavelet transform (DWT) to get three sparse matrices. Next, the obtained matrices are observed by two asymptotical deterministic random measurement matrices based on information entropy and counter (ADMMIC), which not only encrypts the plaintext image, but also compresses it in proportion to reduce the transmission bandwidth and storage space. Subsequently, the corresponding measurement value matrices are shuffled by a double random scrambling based on Arnold map and index vector (DRSAIV) to eliminate the correlation between adjacent pixels. Furthermore, the obtained permutated matrices are diffused by a simultaneous multiple random diffusion of inter–intra components (SMRDIC) to obtain the final cipher image, the plaintext pixel to be diffused, the key matrix involved in diffusion and the position of the obtained ciphertext pixel are all unpredictable, which makes statistical attack invalid. In addition, information entropy values of plaintext image are obtained to generate the initial values of the used chaotic systems, which greatly improve the ability to resist the known-plaintext and chosen-plaintext attacks. Simulation results and security analyses verify that this algorithm has good compression and high security. [ABSTRACT FROM AUTHOR]

In this paper, an effective image encryption algorithm based on 3D Latin cube and memristive hyperchaotic system is proposed. The architecture of preprocessing, permutation and diffusion is utilized. The original image is updated for the first time by the preprocessing process to modify statistical distribution of image pixels, which includes random data embedding and alphabetic XOR operation (AXO). Subsequently, the resulting image is converted to a 3D bit matrix, permutation based on 3D Latin cube (P3DL) and diffusion based on 3D Latin cube (D3DL) are successively manipulated on it to obtain the final cipher image, such that each bit element of plain image may move to any position across the bit plane, and the general transformation of matrix dimensions from permutation to diffusion is also unnecessary. Additionally, the adoption of preprocessing operation makes the permutation easy to resist attacks against special images. Moreover, random data used in preprocessing and 3D Latin cubes utilized in P3DL and D3DL are controlled by a 4D memristive hyperchaotic system, its initial values are computed by information entropy of plain image, which makes the chosen-plaintext and known-plaintext attacks invulnerable. Simulation results and performance analyses verify the effectiveness and security of the proposed image encryption scheme. [ABSTRACT FROM AUTHOR]

A series of four‐armed poly(d,l‐lactide)‐block‐poly(d‐lactide) (4‐DL‐D) copolymers were synthesized by ring‐opening polymerization. By fixing the poly(d,l‐lactide) (PDLLA) block length (1 kg mol−1) and changing the poly(d‐lactide) (PDLA) block length (Mn,PDLA = 0, 0.5, 1.1, 1.3, 1.8 and 2.6 kg mol−1), the crystallization and alkaline degradation of the PLLA/4‐DL‐D blends were investigated. The four‐armed PDLLA core of the copolymer inhibited the crystallization of PLLA, while the outer PDLA block could affect the crystallization differently when its length changed. If Mn,PDLA was 0 or 0.5 kg mol−1, the crystallization of PLLA in the PLLA/4‐DL‐D blend was retarded markedly and the degradation rate of the blend films was much faster than that of neat PLLA film. Interestingly, when Mn,PDLA was 1.1 kg mol−1 or higher, stereocomplex (SC) crystallites with different morphologies were formed, and the degradation rate of the PLLA/4‐DL‐D blend decreased gradually with increasing Mn,PDLA. In the PLLA/4‐DL‐D1.1 blend, the SC crystallites acted as nucleators for PLLA homocrystallites, while in the PLLA/4‐DL‐D1.3 blend, small isolated SC crystallites were observed inside the PLLA homospherulites. When Mn,PDLA was 1.8 or 2.6 kg mol−1, a network structure of SC crystallites was formed and the degradation resistance of the films was markedly enhanced. A possible isothermal crystallization mechanism was proposed for the PLLA/4‐DL‐D blends, and the relationship between the crystallization state and degradation behavior was explored. This work revealed that the crystallization state, which was controlled by the PDLA block length, had a significant effect on the degradation behavior of PLLA/4‐DL‐D blend films. © 2020 Society of Chemical Industry [ABSTRACT FROM AUTHOR]

Organic conjugated materials combining high conductivity with strong solid-state emission are highly desired for organic electronic applications, yet still rather rare. Herein, a novel luminogen (TEN) comprised by linking naphthalene diimides and tri-phenyl ethylene with vinyl bridges is reported. TEN exhibits aggregation-induced emission (AIE) behavior of a strong near-infrared fluorescence over 700 nm and the efficiency above 60.5% in the solid state, while also shows promising application in vivo bio-imaging with good permeability and extremely low background. Single crystal of TEN reveals intra- and intermolecular C-H⋯O hydrogen bonds, contributing to an inclined molecular stacking along the a-axis of the cell, creating a 1D charge carrier transporting channel under a short π-π interaction distance of 3.42 Å, which might benefit the solid emission and charge transport ability simultaneously. Solution processed bottom contact, top gate organic field effect transistors based on TEN reveal a high ambipolar charge transport ability with the hole mobility up to 0.13 cm2 V−1 s−1 and electron mobility up to 0.010 cm2 V−1 s−1. Further atomic force microscopy and X-ray diffraction analysis on TEN thin film confirm the existence of the 1D π-π stacking channel, suggesting the stacking geometry revealed in crystal crucial for facilitating high charge carrier mobility while preserving the strong solid emission at the same time. [ABSTRACT FROM AUTHOR]

At present, information entropies of cipher images gotten by some CS-based image cryptosystems are lower than 7, which make them vulnerable to entropy attack. To cope with this problem, we propose a novel image compression–encryption method based on compressive sensing (CS) and game of life (GOL). Encryption architecture of permutation, compression and diffusion is utilized. Firstly, a plaintext-dependent game-of-life-based scrambling method is presented to shuffle the sparse coefficient matrix of plain image, and the permutation matrix is constructed by rules of GOL, which may effectively reduce the adjacent pixel correlation and enhance the scrambling effect. Secondly, the confused matrix is compressed by CS and diffused using a key matrix to get the cipher image. Additionally, a five-dimensional (5D) memristive hyperchaotic system is used to generate chaotic sequences. They are utilized to construct measurement matrix, to generate initial cell matrix of GOL and to produce key matrix. Information entropy of plain image and external key parameters are combined to compute initial values of the hyperchaotic system. Therefore, our algorithm has high sensitivity to original image and it may resist against known-plaintext attack and chosen-plaintext attack. Experimental results and performance analyses demonstrate that the proposed encryption algorithm is effective to withstand various typical attacks, and it may be applied for image secure communication. [ABSTRACT FROM AUTHOR]

Nowadays, many image cryptosystems have been cracked by chosen-plaintext attacks, for they are not highly sensitive to plain image. To solve this problem, we introduce a plaintext-related mechanism for secure color image encryption, and it is established in the generation and selection of chaotic sequences, permutation and diffusion. In the proposed image cryptosystem, the architecture of permutation and diffusion is adopted. Firstly, plaintext-related Latin-square-based block permutation is proposed to randomly shuffle pixels of the color plain image, diffusion method dependent on the plaintext and scrambled image is further given to modify pixels of permutated image, and finally cipher image is gotten. The chaotic sequence for diffusing the current pixel is dynamically generated according to the plain image and scrambled image, and diffusion operations of red, green and blue components of color plain image affect each other. Besides, chaotic sequences used in encryption are produced by new one-dimensional chaotic systems and dynamically selected, and initial values of chaotic systems are computed by plain image and external keys. Experimental results and security analyses demonstrate that our image encryption has large key space and high security level, and it can be applied for the secure communication of image information. [ABSTRACT FROM AUTHOR]

In this paper, an efficient image compression and encryption scheme combining the parameter-varying chaotic system, elementary cellular automata (ECA) and block compressive sensing (BCS) is presented. The architecture of permutation, compression and re-permutation is adopted. Firstly, the plain image is transformed by DWT, and four block matrices are gotten, and they are a low-frequency block with important information and three high-frequency blocks with less important information. Secondly, ECA is used to scramble the four sparse block matrices, which can effectively change the position of the elements in the matrices and upgrade the confusion effect of the algorithm. Thirdly, according to the importance of each block, BCS is adopted to compress and encrypt four scrambled matrices with different compression ratios. In the BCS, the measurement matrices are constructed by a parameter-varying chaotic system, and thus few parameters may produce the large measurement matrices, which may effectively reduce memory space and transmission bandwidth. Finally, the four compressed matrices are recombined into a large matrix, and the cipher image is obtained by re-scrambling it. Moreover, the initial values of the chaotic system are produced by the SHA 256 hash value of the plain image, which makes the proposed encryption algorithm highly sensitive to the original image. Experimental results and performance analyses demonstrate its good security and robustness. [ABSTRACT FROM AUTHOR]

In this paper, an image encryption algorithm based on 3-D DNA level permutation and substitution scheme is proposed. In order to improve the security level of the image encryption, a 3-D DNA level hybrid permutation scheme using chaotic sequence sorting and 3-D cat map is presented to effectively shuffle the element positions, a novel 3-D DNA level substitution method based on random number is given to directly change the elements of the 3-D DNA matrix, and dynamic and random DNA encoding rules and DNA decoding rules are designed to upgrade the randomness of the encryption results. Moreover, a 6-D hyperchaotic system is used to generate the chaotic sequence for all the encryption processes, and SHA 256 hash value of the plain image is utilized to compute the initial values of chaotic system and system parameters of the 3-D cat map. Simulation results and performance analyses demonstrate that the proposed encryption scheme has large key space, high key sensitivity and may resist against some typical attacks, and it may effectively secure the secret image information. [ABSTRACT FROM AUTHOR]

Medical image encryption may help protect medical privacy. In this paper, we propose a new medical image encryption scheme combined Latin square and chaotic system. The architecture of permutation and diffusion is adopted. Using Latin square and the plain image information, permutation based on plain image and Latin square (PPILS) is presented to shuffle the pixels of the plain image to different rows and columns, effectively weaken the strong correlations between adjacent pixels, and different images have different permutation effect. To improve the encryption effect, bi-directional adaptive diffusion is proposed to spread little change of plain images to the entire pixels of cipher images. Chaotic sequences employed in permutation and diffusion are generated from the four-dimensional memristive chaotic system, its initial values are computed by SHA 256 hash value of the plain image, and thus the proposed algorithm may withstand known-plaintext and chosen-plaintext attacks. Simulation results and performance analyses show that our image encryption scheme has good security and robustness, and it may be applied for medical image encryption applications. [ABSTRACT FROM AUTHOR]

The addition of objects to the original image in splice forgery results in a change in the semantics of the original image, and distribution of these spliced images may bring negative impacts. To solve this issue, many forgery detection methods based on convolutional neural networks are presented. However, they tend to extract deep features, but ignore the importance of shallow semantics. In addition, the complexity of the forgery localization task leads to insufficient accuracy in detecting smaller forged regions. Based on the above problems, an end-to-end image splicing localization network based on multi-scale features and a residual refinement module (RRM) is proposed in this work. Our approach can be roughly divided into two modules: the detection module and the RRM. First, shallow and deep features are extracted from a backbone network, and deep features are then processed by the deeper atrous spatial pyramid pooling (ASPP) module to extract multi-scale features. The deeper ASPP module uses a smaller dilation rate and light convolution, which is more suitable for detecting complex counterfeit images. Second, the shallow features are fused with the multi-scale features to complement the shallow semantic information, such as texture and edges, which further improve the robustness of the model. Finally, the detection network generates coarse prediction maps that are fed to the RRM, and the RRM optimizes these masks by smoothing the boundaries, filling small gaps, and enhancing the edge details, which improves the pixel-level segmentation results for forgery detection. Extensive experimental results on several public datasets show that this method outperforms other state-of-the-art methods in image forgery localization. [ABSTRACT FROM AUTHOR]

Pulsating heat pipes (PHPs) are promising heat transfer devices due to their high heat transfer rate, simple construction, and low manufacturing cost. Most previous investigations focused mainly on their performances at room temperature. In order to have a better understanding on their thermal performances in the cryogenic field, a PHP element with liquid hydrogen as the working fluid has been fabricated. In this paper, the effect of adiabatic length (100 and 500 mm) on the thermal performance of the hydrogen PHP was investigated in order to determine whether it is capable of transferring heat over a long distance. Experiments were carried out at the filling ratio of 51% and the heat loads from 0 to 14 W. It can be observed that the two PHPs are able to maintain stable temperature profiles during the operation as the heat load increases. Their thermal resistances are almost constant around 0.3 K/W, but that of the longer PHP is about 30% larger. The results demonstrate the potential advantages of PHPs to effectively transport heat over a long distance. [ABSTRACT FROM AUTHOR]

In the paper, a novel image encryption algorithm based on DNA sequence operations and chaotic systems is proposed. The encryption architecture of permutation and diffusion is adopted. Firstly, 256-bit hash value of the plain image is gotten to calculate the initial values and system parameters of the 2D Logistic-adjusted-Sine map (2D-LASM) and a new 1D chaotic system; thus, the encryption scheme highly depends on the original image. Next, the chaotic sequences from 2D-LASM are used to produce the DNA encoding/decoding rule matrix, and the plain image is encoded into a DNA matrix according to it. Thirdly, DNA level row permutation and column permutation are performed on the DNA matrix of the original image, inter-DNA-plane permutation and intra-DNA-plane permutation can be attained simultaneously, and then, DNA XOR operation is performed on the permutated DNA matrix using a DNA key matrix, and the key matrix is produced by the combination of two 1D chaotic systems. Finally, after decoding the confused DNA matrix, the cipher image is obtained. Experimental results and security analyses demonstrate that the proposed scheme not only has good encryption effect, but also is secure enough to resist against the known attacks. [ABSTRACT FROM AUTHOR]

Image encryption is an efficient technique for image protection. This paper presents a double color image encryption (DCIE) scheme based on three-dimensional (3D) Brownian motion. The architecture of diffusion, confusion and diffusion is adopted. The proposed algorithm firstly decomposes two color plain images into R, G, B components and bit planes, then conducts XOR operation between every bit plane and a key bit plane generated from carrier images; next, arranges all the bit planes into a 3D bit matrix, performs block-in 3D Brownian motion confusion and block confusion on it, which can permute the elements across bit planes and across plain images; finally, a subsequent diffusion process is utilized to improve the encryption effect. Simulation results and performance analyses illustrate that the proposed encryption scheme can simultaneously encrypt two color images into noise-like ones that have strong ability of resisting various known attacks. [ABSTRACT FROM AUTHOR]

Selective encryption algorithms have emerged as a popular technique for protecting the privacy of images during real-time transmission. For selectively encrypted images, it is necessary to evaluate their security and usability with visual security indices, and there have been a series of studies in this area. However, those proposed visual security indices (VSI) are often ineffective. We present a multi-directional structure and content-aware features-based visual security index (MCVSI) to perform an objective assessment on selectively encrypted images. Considering that selectively encrypted images prevent the main contents from being easily identified, stable local features in the images are extracted to indicate the degree of image content leakage. Meanwhile, we extract spatial structure information that closely aligns with human visual perception to indicate the level of variation in the overall image skeleton. Next, these features are subjected to similarity measurements to produce two types of similarity, content perception feature similarity and structure feature similarity. Finally, our visual security index is built by connecting all feature similarities and their corresponding visual security scores using the regression module. The experimental results and analyses indicate that the proposed MCVSI outperforms many existing mainstream VSI in terms of higher performance and stronger robustness, particularly on low and medium quality images. [ABSTRACT FROM AUTHOR]

Abstract: Fluorescent dyes with multi‐functionality are of great interest for photo‐based cancer theranostics. However, their low singlet oxygen quantum yield impedes their potential applications for photodynamic therapy (PDT). Now, a molecular self‐assembly strategy is presented for a nanodrug with a remarkably enhanced photodynamic effect based on a dye‐chemodrug conjugate. The self‐assembled nanodrug possesses an increased intersystem crossing rate owing to the aggregation of dye, leading to a distinct singlet oxygen quantum yield (Φ(1O2)). Subsequently, upon red light irradiation, the generated singlet oxygen reduces the size of the nanodrug from 90 to 10 nm, which facilitates deep tumor penetration of the nanodrug and release of chemodrug. The nanodrug achieved in situ tumor imaging and potent tumor inhibition by deep chemo‐PDT. Our work verifies a facile and effective self‐assembly strategy to construct nanodrugs with enhanced performance for cancer theranostics. [ABSTRACT FROM AUTHOR]