Generalizing to Out-of-Sample Degradations via Model Reprogramming.

Existing image restoration models are typically designed for specific tasks and struggle to generalize to out-of-sample degradations not encountered during training. While zero-shot methods can address this limitation by fine-tuning model parameters on testing samples, their effectiveness relies on predefined natural priors and physical models of specific degradations. Nevertheless, determining out-of-sample degradations faced in real-world scenarios is always impractical. As a result, it is more desirable to train restoration models with inherent generalization ability. To this end, this work introduces the Out-of-Sample Restoration (OSR) task, which aims to develop restoration models capable of handling out-of-sample degradations. An intuitive solution involves pre-translating out-of-sample degradations to known degradations of restoration models. However, directly translating them in the image space could lead to complex image translation issues. To address this issue, we propose a model reprogramming framework, which translates out-of-sample degradations by quantum mechanic and wave functions. Specifically, input images are decoupled as wave functions of amplitude and phase terms. The translation of out-of-sample degradation is performed by adapting the phase term. Meanwhile, the image content is maintained and enhanced in the amplitude term. By taking these two terms as inputs, restoration models are able to handle out-of-sample degradations without fine-tuning. Through extensive experiments across multiple evaluation cases, we demonstrate the effectiveness and flexibility of our proposed framework. Our codes are available at https://github.com/ddghjikle/Out-of-sample-restoration.

Effects of deep frying and baking on the quality attributes, water distribution, and flavor characteristics of duck jerky.

Introduction: The nutritional value of duck meat is well acknowledged due to its low cholesterol and high protein content. Nevertheless, the impacts of deep-frying and baking on its quality characteristics are not extensively documented in literature. Methods: The objective of this study is to examine the effects of deep-frying, pre-boilingdeep-frying, baking, and pre-boiling-baking on the quality attributes, water distribution, microstructure, and flavor characteristics of duck jerky. Results and discussion: The findings revealed that the deep-frying group had better quality attributes than the baking, pre-boiling-deep-frying, and pre-boiling-baking groups. The deepfried duck jerky had a higher a* value (4.25) and a lower b* value (5.87), with a more appropriate texture profile, and had the highest comprehensive impression score (5.84). Moreover, the drying rate was faster, and the intensity of the free water and oil signal was significantly elevated in the deep-frying group. The microstructure results indicated that the muscle fibers in the deep-frying group were closely packed, whereas those in the baking group were relatively loose. Furthermore, the GC-IMS test revealed that the deep-fried duck jerky had a wider range of volatile flavor compounds, including 11 unique compounds that were only found in this particular product.

Advances in the Preparation, Stability, Metabolism, and Physiological Roles of Anthocyanins: A Review.

Anthocyanins are natural flavonoid polyphenolic compounds widely found in fruits and vegetables. They exhibit antioxidant properties and prophylactic effects in the immune and cardiovascular systems, confer protection against cancer, and contribute to the prevention of cardiovascular diseases. Thus, their incorporation into functional foods, pharmaceuticals, supplements, and cosmetic formulations aims at promoting human well-being. This review comprehensively outlined the structural attributes of anthocyanins, expanding upon diverse methodologies employed for their extraction and production. Additionally, the stability, metabolic pathways, and manifold physiological functions of anthocyanins were discussed. However, their constrained fat solubility, susceptibility to instability, and restricted bioavailability collectively curtail their applicability and therapeutic efficacy. Consequently, a multidimensional approach was imperative, necessitating the exploration of innovative pathways to surmount these limitations, thereby amplifying the utilitarian significance of anthocyanins and furnishing pivotal support for their continual advancement and broader application.

Zein-yeast carboxymethyl glucan particles formed by anti-solvent precipitation for encapsulating resveratrol.

In the work, zein-yeast carboxymethyl glucan (ZY) particles were fabricated by a novel ultrasonic assisted anti-solvent precipitation (ASP) method, which was a good delivery system for resveratrol. The particle size and zeta-potential of ZY samples were detected by Zetasizer Pro analyzer, they gradually increased as the mass ratio of zein and yeast carboxymethyl glucan (YCG) changed from 10:1 to 10:5. The intermolecular interactions were investigated by zeta-potentiometric analyzer, Fourier transform infrared spectroscopy and fluorescence spectroscopy. Electrostatic interaction, hydrogen bonding and hydrophobic effects between zein and YCG molecules were identified as the main driving forces in the formation of ZY particles. The optimized ZY (10:3) binary particles were used as delivery system for encapsulating and protecting resveratrol. They had high encapsulation efficiency (85.4 %) and loading capacity (6.1 %), and increased the retention rate of resveratrol by 2.10 and 1.21 folds after exposure to light and heat conditions, effectively protect resveratrol against light and thermal degradation. These particles also delayed the release of resveratrol in simulated gastrointestinal digestion, which might improve its oral bioavailability. In conclusion, ZY binary particles could be regarded as a useful and promising delivery vehicle, which might contribute to the application of hydrophobic bioactive ingredients in functional foods.

Research Advances in Plant Protein-Based Products: Protein Sources, Processing Technology, and Food Applications.

Plant proteins are high-quality dietary components of food products. With the growing interest in sustainable and healthy food alternatives, plant proteins have gained significant attention as viable substitutes for animal-based proteins. Understanding the diversity of protein sources derived from plants, novel processing technology, and multiple applications is crucial for developing nutritious and sustainable plant protein-based products. This Review summarizes the natural sources of traditional and emerging plant proteins. The classifications, processing technologies, and applications of plant protein-based products in the food industry are explicitly elucidated. Moreover, the advantages and disadvantages of plant protein-based food products are revealed. Strategies such as protein fortification and complementation to overcome these shortcomings are critically discussed. We also demonstrate several issues that need to be addressed in future development.

Camouflaged Object Segmentation Based on Matching-Recognition-Refinement Network.

In the biosphere, camouflaged objects take the advantage of visional wholeness by keeping the color and texture of the objects highly consistent with the background, thereby confusing the visual mechanism of other creatures and achieving a concealed effect. This is also the main reason why the task of camouflaged object detection is challenging. In this article, we break the visual wholeness and see through the camouflage from the perspective of matching the appropriate field of view. We propose a matching-recognition-refinement network (MRR-Net), which consists of two key modules, i.e., the visual field matching and recognition module (VFMRM) and the stepwise refinement module (SWRM). In the VFMRM, various feature receptive fields are used to match candidate areas of camouflaged objects of different sizes and shapes and adaptively activate and recognize the approximate area of the real camouflaged object. The SWRM then uses the features extracted by the backbone to gradually refine the camouflaged region obtained by VFMRM, thus yielding the complete camouflaged object. In addition, a more efficient deep supervision method is exploited, making the features from the backbone input into the SWRM more critical and not redundant. Extensive experimental results demonstrate that our MRR-Net runs in real-time (82.6 frames/s) and significantly outperforms 30 state-of-the-art models on three challenging datasets under three standard metrics. Furthermore, MRR-Net is applied to four downstream tasks of camouflaged object segmentation (COS), and the results validate its practical application value. Our code is publicly available at: https://github.com/XinyuYanTJU/MRR-Net.

Path-enhanced graph convolutional networks for node classification without features.

Most current graph neural networks (GNNs) are designed from the view of methodology and rarely consider the inherent characters of graph. Although the inherent characters may impact the performance of GNNs, very few methods are proposed to resolve the issue. In this work, we mainly focus on improving the performance of graph convolutional networks (GCNs) on the graphs without node features. In order to resolve the issue, we propose a method called t-hopGCN to describe t-hop neighbors by the shortest path between two nodes, then the adjacency matrix of t-hop neighbors as features to perform node classification. Experimental results show that t-hopGCN can significantly improve the performance of node classification in the graphs without node features. More importantly, adding the adjacency matrix of t-hop neighbors can improve the performance of existing popular GNNs on node classification.

Improvement of physicochemical stability and digestive properties of quercetagetin using zein-chondroitin sulfate particles prepared by antisolvent co-precipitation.

Zein-quercetagetin-chondroitin sulfate (Zein-Que-CS) composite nanoparticles with different compositions were successfully fabricated using a novel antisolvent co-precipitation method. The mean particle diameter (97.5 to 219.4 nm), negative surface potential (-29.9 to -51.1 mV), and turbidity (265 to 370 NTU) of suspensions of Zein-Que nanoparticles increased after the addition of CS. Electrostatic attraction, hydrogen bonding, and hydrophobic attraction were the main driving forces for the formation of the composite nanoparticles. The encapsulation efficiency and loading capacity of the quercetagetin within the Zein-Que-CS (100:10:30) nanoparticles were 91.6 % and 6.1 %, respectively. The photostability and thermal stability of the encapsulated quercetagetin were 3.4- and 3.2- fold higher than that of the free form. The nanoparticles had good resistance to sedimentation and exhibited slow-release properties under simulated gastrointestinal conditions. The Zein-Que-CS nanoparticles developed in this study may therefore be useful for the encapsulation, protection, and delivery of quercetagetin.

Insight into the roles of microalgae on simultaneous nitrification and denitrification in microalgal-bacterial sequencing batch reactors: Nitrogen removal, extracellular polymeric substances, and microbial communities.

This study explored the influence and mechanism of microalgae on simultaneous nitrification and denitrification (SND) in microalgal-bacterial sequencing batch reactors (MB-SBR). It particularly focused on nitrogen transformation in extracellular polymeric substances (EPS) and functional groups associated with nitrogen removal. The results showed that MB-SBR achieved more optimal performance than control, with an SND efficiency of 68.01% and total nitrogen removal efficiency of 66.74%. Further analyses revealed that microalgae changed compositions and properties of EPS by increasing EPS contents and improving transfer, conversion, and storage capacity of nitrogen in EPS. Microbial community analysis demonstrated that microalgae promoted the enrichment of functional groups and genes related to SND and introduced diverse nitrogen removal pathways. Moreover, co-occurrence network analysis elucidated the interactions between communities of bacteria and microalgae and the promotion of SND by microalgae as keystone connectors in the MB-SBR. This study provides insights into the roles of microalgae for enhanced SND.

Query-Efficient Black-Box Adversarial Attack With Customized Iteration and Sampling.

It is a challenging task to fool an image classifier based on deep neural networks under the black-box setting where the target model can only be queried. Among existing black-box attacks, transfer-based methods tend to overfit the substitute model on parameter settings. Decision-based methods have low query efficiency due to fixed sampling and greedy search strategy. To alleviate the above problems, we present a new framework for query-efficient black-box adversarial attack by bridging transfer-based and decision-based attacks. We reveal the relationship between current noise and variance of sampling, the monotonicity of noise compression, and the influence of transition function on the decision-based attack. Guided by the new framework, we propose a black-box adversarial attack named Customized Iteration and Sampling Attack (CISA). CISA estimates the distance from nearby decision boundary to set the stepsize, and uses a dual-direction iterative trajectory to find the intermediate adversarial example. Based on the intermediate adversarial example, CISA conducts customized sampling according to the noise sensitivity of each pixel to further compress noise, and relaxes the state transition function to achieve higher query efficiency. Extensive experiments demonstrate CISA's advantage in query efficiency of black-box adversarial attacks.

Investigation of a novel smart and active packaging materials: Nanoparticle-filled carrageenan-based composite films.

In this study, curcumin, zein, epigallocatechin gallate (EGCG) and carrageenan were used to fabricate curcumin-zein-EGCG-carrageenan (CZEC) layer-by-layer nanoparticles. These nanoparticles were embedded within carrageenan films to form composite films with enhanced antioxidant activity and smart responses. Light scattering, micro-electrophoresis, FTIR, and SEM were used to characterize the size, charge, interactions, and morphology of CZEC nanoparticles. The optical, mechanical, morphological, spectroscopic, thermal, and other functional attributes of the films were evaluated. The CZEC nanoparticles were uniformly dispersed within the carrageenan matrices, and improved their UV barrier (2.4-11.1 A mm-1), mechanical (7.09 %-9.35 %), and thermal resistance properties. The films exhibited a color change, from yellow to red, in response to an increase in pH (2.0-12.0) or ammonia concentration (8.0 mM). The films also displayed relatively high DPPH (79.46 %) and ABTS (73.34 %) free radical scavenging activities. Finally, the composite film exhibited the ability of monitoring and extending the freshness of packaged fish.

Melanin/melanin-like nanoparticles: As a naturally active platform for imaging-guided disease therapy.

The development of biocompatible and efficient nanoplatforms that combine diagnostic and therapeutic functions is of great importance for precise disease treatment. Melanin, an endogenous biopolymer present in living organisms, has attracted increasing attention as a versatile bioinspired functional platform owing to its unique physicochemical properties (e.g., high biocompatibility, strong chelation of metal ions, broadband light absorption, high drug binding properties) and inherent antioxidant, photoprotective, anti-inflammatory, and anti-tumor effects. In this review, the fundamental physicochemical properties and preparation methods of natural melanin and melanin-like nanoparticles were outlined. A systematical description of the recent progress of melanin and melanin-like nanoparticles in single, dual-, and tri-multimodal imaging-guided the visual administration and treatment of osteoarthritis, acute liver injury, acute kidney injury, acute lung injury, brain injury, periodontitis, iron overload, etc. Was then given. Finally, it concluded with a reasoned discussion of current challenges toward clinical translation and future striving directions. Therefore, this comprehensive review provides insight into the current status of melanin and melanin-like nanoparticles research and is expected to optimize the design of novel melanin-based therapeutic platforms and further clinical translation.

Curiosity-Driven Salient Object Detection With Fragment Attention.

Recent deep learning based salient object detection methods with attention mechanisms have made great success. However, existing attention mechanisms can be generally separated into two categories. One part chooses to calculate weights indiscriminately, which yields computational redundancy. While one part focuses randomly on a small part of the images, such as hard attention, resulting in incorrectness owing to insufficiently targeted selection of a subset of tokens. To alleviate these problems, we design a Curiosity-driven Network (CNet) and a Curiosity-driven Learning Algorithm (CLA) based on fragment attention (FA) mechanism newly defined in this paper. FA imitates the process of cognition perception driven by human curiosity, and divides the degree of curiosity into three levels, i.e. curious, a little curious and not curious. These three levels correspond to five saliency degrees, including salient and non-salient, likewise salient and likewise non-salient, completely uncertain. With more knowledge gained by the network, CLA transforms the curiosity degree of each pixel to yield enhanced detail-enriched saliency maps. In order to extract more context-aware information of potential salient objects and make a better foundation for CLA, a high-level feature extraction module (HFEM) is further proposed. Based on the much better high-level features extracted by HFEM, FA can classify the curiosity degree for each pixel more reasonably and accurately. Extensive experiments on five popular datasets clearly demonstrate that our method outperforms the state-of-the-art approaches without any pre-processing operations or post-processing operations.

Recent advances in carrageenan-based films for food packaging applications.

In order to solve the increasingly serious environmental problems caused by plastic-based packaging, carrageenan-based films are drawing much attentions in food packaging applications, due to low cost, biodegradability, compatibility, and film-forming property. The purpose of this article is to present a comprehensive review of recent developments in carrageenan-based films, including fabrication strategies, physical and chemical properties and novel food packaging applications. Carrageenan can be extracted from red algae mainly by hydrolysis, ultrasonic-assisted and microwave-assisted extraction, and the combination of multiple extraction methods will be future trends in carrageenan extraction methods. Carrageenan can form homogeneous film-forming solutions and fabricate films mainly by direct coating, solvent casting and electrospinning, and mechanism of film formation was discussed in detail. Due to the inherent limitations of the pure carrageenan film, physical and chemical properties of carrageenan films were enhanced by incorporation with other compounds. Therefore, carrageenan-based films can be widely used for extending the shelf life of food and monitoring the food freshness by inhibiting microbial growth, reducing moisture loss and the respiration, etc. This article will provide useful guidelines for further research on carrageenan-based films.

Action Keypoint Network for Efficient Video Recognition.

Reducing redundancy is crucial for improving the efficiency of video recognition models. An effective approach is to select informative content from the holistic video, yielding a popular family of dynamic video recognition methods. However, existing dynamic methods focus on either temporal or spatial selection independently while neglecting a reality that the redundancies are usually spatial and temporal, simultaneously. Moreover, their selected content is usually cropped with fixed shapes (e.g., temporally-cropped frames, spatially-cropped patches), while the realistic distribution of informative content can be much more diverse. With these two insights, this paper proposes to integrate temporal and spatial selection into an Action Keypoint Network (AK-Net). From different frames and positions, AK-Net selects some informative points scattered in arbitrary-shaped regions as a set of "action keypoints" and then transforms the video recognition into point cloud classification. More concretely, AK-Net has two steps, i.e., the keypoint selection and the point cloud classification. First, it inputs the video into a baseline network and outputs a feature map from an intermediate layer. We view each pixel on this feature map as a spatial-temporal point and select some informative keypoints using self-attention. Second, AK-Net devises a ranking criterion to arrange the keypoints into an ordered 1D sequence. Since the video is represented with a 1D sequence after the specified layer, AK-Net transforms the subsequent layers into a point cloud classification sub-net by compacting the original 2D convolutional kernels into 1D kernels. Consequentially, AK-Net brings two-fold benefits for efficiency: The keypoint selection step collects informative content within arbitrary shapes and increases the efficiency for modeling spatial-temporal dependencies, while the point cloud classification step further reduces the computational cost by compacting the convolutional kernels. Experimental results show that AK-Net can consistently improve the efficiency and performance of baseline methods on several video recognition benchmarks.

Advances in Protein-Based Nanocarriers of Bioactive Compounds: From Microscopic Molecular Principles to Macroscopical Structural and Functional Attributes.

Many proteins can be used to fabricate nanocarriers for encapsulation, protection, and controlled release of nutraceuticals. This review examined the protein-based nanocarriers from microscopic molecular characteristics to the macroscopical structural and functional attributes. Structural, physical, and chemical properties of protein-based nanocarriers were introduced in detail. The spatial size, shape, water dispersibility, colloidal stability, etc. of protein-based nanocarriers were largely determined by the molecular physicochemical principles of protein. Different preparative techniques, including antisolvent precipitation, pH-driven, electrospray, and gelation methods, among others, can be used to fabricate different protein-based nanocarriers. Various modifications based on physical, chemical, and enzymatic approaches can be used to improve the functional performance of these nanocarriers. Protein is a natural resource with a wide range of sources, including plant, animal, and microbial, which are usually used to fabricate the nanocarriers. Protein-based nanocarriers have many advantages in aid of the application of bioactive ingredients to the medical, food, and cosmetic industries.

Production, Characterization, Delivery, and Cholesterol-Lowering Mechanism of Phytosterols: A Review.

Phytosterols are natural plant-based bioactive compounds that can lower blood cholesterol levels and help prevent cardiovascular diseases. Consequently, they are being utilized in functional foods, supplements, and pharmaceutical products designed to improve human health. This paper summarizes different approaches to isolate, purify, and characterize phytosterols. It also discusses the hypolipidemic mechanisms of phytosterols and their impact on cholesterol transportation. Phytosterols have a low water-solubility, poor chemical stability, and limited bioavailability, which limits their utilization and efficacy in functional foods. Strategies are therefore being developed to overcome these shortcomings. Colloidal delivery systems, such as emulsions, oleogels, liposomes, and nanoparticles, have been shown to be effective at improving the water-dispersibility, stability, and bioavailability of phytosterols. These delivery systems can be used to incorporate phytosterols into a broader range of cholesterol-lowering functional foods and beverages. We also discuses several issues that need to be addressed before these phytosterol delivery systems can find widespread commercial utilization.

Instance-Invariant Domain Adaptive Object Detection Via Progressive Disentanglement.

Most state-of-the-art methods of object detection suffer from poor generalization ability when the training and test data are from different domains. To address this problem, previous methods mainly explore to align distribution between source and target domains, which may neglect the impact of the domain-specific information existing in the aligned features. Besides, when transferring detection ability across different domains, it is important to extract the instance-level features that are domain-invariant. To this end, we explore to extract instance-invariant features by disentangling the domain-invariant features from the domain-specific features. Particularly, a progressive disentangled mechanism is proposed to decompose domain-invariant and domain-specific features, which consists of a base disentangled layer and a progressive disentangled layer. Then, with the help of Region Proposal Network (RPN), the instance-invariant features are extracted based on the output of the progressive disentangled layer. Finally, to enhance the disentangled ability, we design a detached optimization to train our model in an end-to-end fashion. Experimental results on four domain-shift scenes show our method is separately 2.3, 3.6, 4.0, and 2.0 percent higher than the baseline method. Meanwhile, visualization analysis demonstrates that our model owns well disentangled ability.

A Polymer Multicellular Nanoengager for Synergistic NIR-II Photothermal Immunotherapy.

Cell-membrane-coated nanoparticles (CCNPs) that integrate the biophysiological advantages of cell membranes with the multifunctionalities of synthetic materials hold great promise in cancer immunotherapy. However, strategies have yet to be revealed to further improve their immunotherapeutic efficacy. Herein, a polymer multicellular nanoengager (SPNE) for synergistic second-near-infrared-window (NIR-II) photothermal immunotherapy is reported. The nanoengager consists of an NIR-II absorbing polymer as the photothermal core, which is camouflaged with fused membranes derived from immunologically engineered tumor cells and dendritic cells (DCs) as the cancer vaccine shell. In association with the high accumulation in lymph nodes and tumors, the multicellular engagement ability of the SPNE enables effective cross-interactions among tumor cells, DCs, and T cells, leading to augmented T cell activation relative to bare or tumor-cell-coated nanoparticles. Upon deep-tissue penetrating NIR-II photoirradiation, SPNE eradicates the tumor and induces immunogenic cell death, further eliciting anti-tumor T cell immunity. Such a synergistic photothermal immunotherapeutic effect eventually inhibits tumor growth, prevents metastasis and procures immunological memory. Thus, this study presents a general cell-membrane-coating approach to develop photo-immunotherapeutic agents for cancer therapy.