Range Of Applications Research Articles

In-Domain GAN Inversion for Faithful Reconstruction and Editability.

Generative Adversarial Networks (GANs) have significantly advanced image synthesis through mapping randomly sampled latent codes to high-fidelity synthesized images. However, applying well-trained GANs to real image editing remains challenging. A common solution is to find an approximate latent code that can adequately recover the input image to edit, which is also known as GAN inversion. To invert a GAN model, prior works typically focus on reconstructing the target image at the pixel level, yet few studies are conducted on whether the inverted result can well support manipulation at the semantic level. This work fills in this gap by proposing in-domain GAN inversion, which consists of a domain-guided encoder and a domain-regularized optimizer, to regularize the inverted code in the native latent space of the pre-trained GAN model. In this way, we manage to sufficiently reuse the knowledge learned by GANs for image reconstruction, facilitating a wide range of editing applications without any retraining. We further make comprehensive analyses on the effects of the encoder structure, the starting inversion point, as well as the inversion parameter space, and observe the trade-off between the reconstruction quality and the editing property. Such a trade-off sheds light on how a GAN model represents an image with various semantics encoded in the learned latent distribution.

Reconstructing Randomly Masked Spectra Helps DNNs Identify Discriminant Wavenumbers.

Nondestructive detection methods, based on vibrational spectroscopy, are vitally important in a wide range of applications including industrial chemistry, pharmacy and national defense. Recently, deep learning has been introduced into vibrational spectroscopy showing great potential. Different from images, text, etc. that offer large labeled data sets, vibrational spectroscopic data is very limited, which requires novel concepts beyond transfer and meta learning. To tackle this, we propose a task-enhanced augmentation network (TeaNet). The key component of TeaNet is a reconstruction module that inputs randomly masked spectra and outputs reconstructed samples that are similar to the original ones, but include additional variations learned from the domain. These augmented samples are used to train the classification model. The reconstruction and prediction parts are trained simultaneously, end-to-end with back-propagation. Results on both synthetic and real-world datasets verified the superiority of the proposed method. In the most difficult synthetic scenarios TeaNet outperformed CNN by 17%. We visualized and analysed the neuron responses of TeaNet and CNN, and found that TeaNet's ability to identify discriminant wavenumbers was excellent compared to CNN. Our approach is general and can be easily adapted to other domains, offering a solution to more accurate and interpretable few-shot learning.

Growing Like a Tree: Finding Trunks From Graph Skeleton Trees.

The message-passing paradigm has served as the foundation of graph neural networks (GNNs) for years, making them achieve great success in a wide range of applications. Despite its elegance, this paradigm presents several unexpected challenges for graph-level tasks, such as the long-range problem, information bottleneck, over-squashing phenomenon, and limited expressivity. In this study, we aim to overcome these major challenges and break the conventional "node- and edge-centric" mindset in graph-level tasks. To this end, we provide an in-depth theoretical analysis of the causes of the information bottleneck from the perspective of information influence. Building on the theoretical results, we offer unique insights to break this bottleneck and suggest extracting a skeleton tree from the original graph, followed by propagating information in a distinctive manner on this tree. Drawing inspiration from natural trees, we further propose to find trunks from graph skeleton trees to create powerful graph representations and develop the corresponding framework for graph-level tasks. Extensive experiments on multiple real-world datasets demonstrate the superiority of our model. Comprehensive experimental analyses further highlight its capability of capturing long-range dependencies and alleviating the over-squashing problem, thereby providing novel insights into graph-level tasks.

Manufacturing and Applications of Cellulosic Films in Packaging: An Alternative for Plastic Films

Background: Different synthetic materials are constantly being utilized as raw materials for packaging, which might have several drawbacks, like plastic, some of which have poor lifespans. Plastic can get damaged, scratched, or fractured over time, easily take in tastes and fragrances, and significantly affect the environment and health. Excessive plastic waste may lead to environmental pollution and related consequences. Plastics that contaminate food have also been linked to health problems like high blood pressure, heart palpitations, headaches, nausea, diarrhea, and flushing. Objective: The present review entails the hazards associated with the usage of plastic as a packaging material and highlights the necessity to reduce its use and promote alternatives and sustainable production. Thus, the present review portrays the need for biopolymers, types of plant polymers, methods of cellulose extraction and derivatization, the manufacturing process of cellulosic films and thereby use as film and coatings for food preservation. The current review also summarizes different quality control tests for effective utilization as an alternative to synthetic polymers. Conclusion: Different polymers derived from agricultural products are excellent candidates for creating packaging materials. Similar to polysaccharides, these polymers, including nanocomposites, films, composites, etc., are helpful in a wide range of applications. Cellulose, one of the most widespread polysaccharides in the world, offers a variety of appealing special qualities. It appears reasonable and must be considered that cellulose can be used effectively as a raw material for packaging. In this review, we explore how cellulose is used in the packaging industry to make composites, coating materials, and edible and non-edible films.

Uncertainty budget for detector-based absolute radiometric calibration with GLAMR

The accuracy of the absolute radiometric calibration (RadCal) for remote sensing instruments is essential to their wide range of applications. The uncertainty associated with the traditional source-based RadCal method is assessed at a 2% (k=1) or higher level for radiance measurement. To further improve the accuracy to meet the demands of climate studies, a detector-based approach using tunable lasers as a light source has been devised. The Goddard Laser for Absolute Measurement of Radiance, known as the GLAMR system, is a notable example of the incorporation of such technology. Using transfer radiometers calibrated at the National Institute of Standards and Technology as calibration standards, the absolute spectral response function of a remote sensing instrument is measured with its uncertainty traceable to the International System of Units. This paper presents a comprehensive uncertainty analysis of the detector-based absolute RadCal using the GLAMR system. It identifies and examines uncertainty sources during the GLAMR RadCal test, including those from the GLAMR system, the testing configuration, and data processing methodologies. Analysis is carried out to quantify the contribution of each source and emphasize the most influential factors. It is shown that the calibration uncertainty of GLAMR RadCal can be better than 0.3% (k=1) in the wavelength range of 350–950 nm and 0.6% (k=1) between 950 and 2300 nm, with the exception of regions with strong water absorption. In addition, recommendations are made to refine the calibration process to further reduce the uncertainty.

Endowing Metal-Organic Coordination Materials with Chiroptical Activity by a Chiral Anion Strategy.

Recently, chiral metal-organic coordination materials have emerged as promising candidates for a wide range of applications in chiroptoelectronics, chiral catalysis, and information encryption, etc. Notably, the chiroptical effect of coordination chromophores makes them appealing for applications such as photodetectors, OLEDs, 3D displays, and bioimaging. The direct synthesis of chiral coordination materials using chiral organic ligands or complexes with metal-centered chirality is very often tedious and costly. In the case of ionic coordination materials, the combination of chiral anions with cationic, achiral coordination compounds through noncovalent interactions may endow molecular materials with desirable chiroptical properties. The use of such a simple chiral strategy has been proven effective in inducing promising circular dichroism and/or circularly polarized luminescence signals. This concept article mainly delves into the latest advances in exploring the efficacy of such a chiral anion strategy for transforming achiral coordination materials into chromophores with superb photo- or electro-chiroptical properties. In particular, ionic small-molecular metal complexes, metal clusters, coordination supramolecular assemblies, and metal-organic frameworks containing chiral anions are discussed. A perspective on the future opportunities on the preparation of chiroptical materials with the chiral anion strategy is also presented.

A Comprehensive Study on Model Simplification and Parameter Estimation of Brushless Doubly Fed Standalone Generation System

Brushless doubly-fed induction generator (BDFIG) has great potential in stand-alone generation applications. Currently, lots of control strategies have been proposed based on the BDFIG d-q model with different machine parameters ignored for simplification, but the simplification influences on model accuracy have not been discussed systematically, and the application ranges are not clear. Therefore, aiming at controller design, this paper provides a comprehensive study on model simplification and parameter estimation of brushless doubly-fed stand- alone generation. Based on the intuitive DFIG-like BDFIG model and the typical electrical quantity relations widely used in stand-alone controller design, steady-state and dynamic parameter sensitivities of BDFIG are analyzed with operation condition fully considered. Based on the analysis, the universal BDFIG simplified d-q model aiming at stand-alone control is evaluated sufficiently, and the application range is clarified. The simplification of control analysis is illustrated, and a highly-targeted and simple parameter estimation method is proposed for simplified modeling and control. Finally, simulations and experiments verify the analysis accuracy of the simplified DFIG-like BDFIG model and the feasibility of the proposed parameter estimation method.

Hybrid Wi-Fi and PLC network for efficient e-health communication in hospitals: a prototype

E-health is being adapted in modern hospitals as a significant addition to the existing healthcare services. To this end, modern hospitals urgently require a mobile, high-capacity, secure, and cost-effective communication infrastructure. In this paper, we explore potential applications of a hybrid broadband power line communication (PLC) and Wi-Fi in an indoor hospital scenario. It utilizes the existing power line cables and Wi-Fi plug-and-play devices for indoor broadband communication. Broadband power line (BPL) adaptors with Wi-Fi outputs are used to build an access network in hospitals, particularly in areas where the wireless router signal is poor. The Tenda PH10 AV1,000 AC Wi-Fi power line adapter is a set of BPL adapters that offer operational bandwidth of up to 1,000 Mbps. These adapters are based on the HomePlug AV2 protocol and can provide a data rate up to 200 Mbps on the physical layer. An experiment using the PLC Wi-Fi kit is carried out to show that a Wi-Fi and PLC hybrid network is the best candidate to provide wide range of practical applications in a hospital including, but not limited to, telemedicine, electronic medical records, early-stage disease diagnosis, health management, real-time monitoring, and remote surgeries.

Parthenolide induces gallbladder cancer cell apoptosis via MAPK signalling.

Parthenolide (PTL) has a wide range of clinical applications owing to its anti-inflammatory and antitumor effects. To date, the antitumor effect of PTL on gallbladder cancer (GBC) remains largely unknown. Therefore, we aimed to investigate the biological effects of PTL on GBC. The cellular viability and proliferation of GBC-SD and NOZ cell lines after treatment with different concentrations of PTL were analyzed using the Cell Counting Kit-8 (CCK8)assay and colony formation assay. Apoptosis analysis was performed using flow cytometry. Hoechst staining was performed. RNA sequencing (RNA-seq) was performed to identify PTL-related genes and signalling pathways. Furthermore, we confirmed the involvement of these signalling pathways by qRT-PCR and western blotting. For the in-vivo experiments, a xenograft model was used to evaluate the effects of PTL on the proliferation of NOZ cells. PTL significantly inhibited GBC cell growth in vitro and induced apoptosis in the GBC-SD and NOZ cell lines in a dose-dependent manner. RNA sequencing data showed that the immune response and mitogen-activated protein kinase (MAPK) signalling pathways are closely associated with PTL-induced gallbladder cancer cell apoptosis. PTL upregulated BAX, cleaved PARP-1, cleaved caspase-3, cleaved caspase-9, P53 and decreased the expression of BCL-2, phosphorylated ERK, and phosphorylated MEK in vitro. Tumour volume and weight were also suppressed by PTL in vivo. Moreover, the effects of PTL on GBC cells might be mediated by the MAPK pathway. PTL significantly inhibits gallbladder cancer cell proliferation and induces apoptosis through the MAPK pathway, which is a potential molecular reagent for treating GBC. However, further exploration is needed to verify the antitumor effects of PTL and its intracellular signalling mechanism.

Structure-Aware Prototypical Neural Process for Few-Shot Graph Classification.

Graph classification plays an important role in a wide range of applications from biological prediction to social analysis. Traditional graph classification models built on graph kernels are hampered by the challenge of poor generalization as they are heavily dependent on the dedicated design of handcrafted features. Recently, graph neural networks (GNNs) become a new class of tools for analyzing graph data and have achieved promising performance. However, it is necessary to collect a large number of labeled graph data for training an accurate GNN, which is often unaffordable in real-world applications. Therefore, it is an open question to build GNNs under the condition of few-shot learning where only a few labeled graphs are available. In this article, we introduce a new Structure-aware Prototypical Neural Process (SPNP for short) for a few-shot graph classification. Specifically, at the encoding stage, SPNP first employs GNNs to capture graph structure information. Then, SPNP incorporates such structural priors into the latent path and the deterministic path for representing stochastic processes. At the decoding stage, SPNP uses a new prototypical decoder to define a metric space where unseen graphs can be predicted effectively. The proposed decoder, which contains a self-attention mechanism to learn the intraclass dependence between graphs, can enhance the class-level representations, especially for new classes. Furthermore, benefited from such a flexible encoding-decoding architecture, SPNP can directly map the context samples to a predictive distribution without any complicated operations used in previous methods. Extensive experiments demonstrate that SPNP achieves consistent and significant improvements over state-of-the-art methods. Further discussions are provided toward model efficiency and more detailed analysis.

SWT-PCA-CNN: hyperspectral image classification with multi-stage feature extraction and parameter tuning

Hyperspectral imaging is an increasingly popular technique in remote sensing, offering a wealth of spectral information for a range of applications. This paper presents a comparative study of hyperspectral image classification techniques using three different datasets: Indian Pines, Salinas, and Pavia University. The study employs a combination of three methods, namely stationary wavelet transforms (SWT), principal component analysis (PCA), and convolutional neural network (CNN), to develop a model for hyperspectral image classification. The proposed approach combines SWT and PCA for spatial feature extraction and dimensionality reduction, followed by classification using CNN. Furthermore, the study performs parameter tuning by changing the optimizer, activation function, and filter size of the CNN model on the Indian Pines dataset. The results demonstrate that the proposed SWT-PCA-CNN approach outperforms the conventional DWT-PCA and PCA-KNN algorithms, achieving an overall classification accuracy of 98.2%, 99.86%, 99.80% on the Indian Pines, Salinas and Pavia University datasets respectively. The study highlights the effectiveness of the proposed approaches for hyperspectral image classification and their potential for applications in remote sensing and other fields.

Double-graviton production from Standard Model plasma

The thermal plasma filling the early universe generated a stochasticgravitational wave background that peaks in the microwave frequencyrange today. If the graviton production rate is expressed as a series in afine-structure constant, α, and the temperature over the Planck mass, T 2/m pl 2, then the lowest-order contributions come from single(∼αT 2/m pl 2) and double (∼T 4/m pl 4)graviton production via 2 → 2 scatterings.We show that in the Standard Model,single-graviton production dominates if the maximaltemperature is smaller than4 × 1018 GeV. This justifies previous calculationswhich relied solely on single-graviton production.We mention Beyond the Standard Model scenarios in which the single anddouble-graviton contributions could be of comparable magnitudes.Finally, we elaborate on what these results imply for the range ofapplicability of General Relativity as an effective theory.

Implementation and measurement of quantum entanglement using IBM quantum platforms

The use of quantum entanglement has garnered increasing attention among researchers in recent years due to its wide range of applications, not only revolutionizing the field of information processing but also enhancing quantum-safe communications. Identifying the degree of entanglement present in quantum states is a crucial focus, and designing an algorithm capable of feasibly measuring entanglement is imperative. While theoretical calculations hold high regard, the ease of implementing these algorithms in a laboratory setting is essential to gauge their efficiency.In this context, IBM quantum computers stand out as discrete value NISQ (Noisy Intermediate-Scale Quantum) platforms These platforms are based on superconducting qubits, providing an opportunity to test our algorithms without the need for extravagant laboratory equipment. This paper proposes an algorithm designed to measure entanglement in a bipartite system. We will execute the algorithm on IBM’s 127-qubit backends to compare our calculations with real-world results. Furthermore, we aim to address and mitigate errors inherent in these devices by utilizing local mitigation technique available in the IBM Experiments Python package, aiming for more accurate and reliable outcomes.

Energy and exergy analysis of the cryogenic system for HTS magnets based on an open-cycle G-M/J-T cryocooler

High-field magnets have been developed for a broad range of research applications. High-temperature superconducting magnets have received considerable attention due to their potential use in high-field applications. The High Magnetic Field Laboratory of the Chinese Academy of Sciences (CHMFL) is currently designing and constructing high-temperature superconducting (HTS) magnets which need the cooling capacity about 5 W to 10 W at 4.5 K in order to ensure the stability of the system. To meet their operational requirements, a miniature cryogenic system based on the Gifford-McMahon/Joule-Thomson (G-M/J-T) cryocooler has been designed and analyzed. In this research, the performance of the cryogenic system for HTS magnets is investigated systematically by employing both energy analysis and entropy analysis techniques. The results highlight the importance of several key factors, including high pressures, precooling temperature at cooling stages, and J-T exchanger efficiency, for improving the cooling capacity of the cryogenic system. Lowering the precooling temperature can enhance the system cooling capacity and also cause an increase in the precooling capacity. As the pressure increases, the cooling capacity of the system reaches its maximum point at 11.31 bar The J-T heat exchanger accounts for the largest exergy loss rate, 35.2–43.7%. The findings provide technical guidelines for the subsequent experiments and performance optimization of cryogenic system.

Rheology of Superabsorbent Polymer-Modified and Basalt Fiber-Reinforced Cement Paste with Silica Fume: Response Surface Methodology

A composite's rheology can be changed by adding superabsorbent polymer (SAP) and basalt fibers and using silica fume. This study aimed to investigate the effects of these components on the viscosity and shear stress parameters of the paste. The proportions of the components were varied, with SAP content ranging from 0.01% to 0.03%, basalt fiber from 0% to 0.50%, silica fume (micro silica) at 15%, and water content from 0.40 to 0.50. Response surface methodology was used to optimize the mixture proportions, and the rheological properties of the resulting pastes were characterized using a rheometer. Results showed that the addition of SAP and basalt fiber had a significant impact on the rheological properties of the paste, with increasing amounts of both resulting in increased viscosity and shear stress. Overall, this study highlights the potential of SAP and basalt fiber in advances of the rheology of cement paste and provides insight into the optimal proportions of these components for achieving desired rheological properties. The findings of this study could be useful in developing high-performance concrete with enhanced rheological properties, which could have a wide range of applications in the construction industry. In addition, 0.50% BF, 0.01% SAP, and 0.445 water-to-cement were found as optimum proportions regarding the rheology of the cement paste.

Catalytic Mechanism and Heterologous Biosynthesis Application of Sesquiterpene Synthases.

Sesquiterpenes comprise a diverse group of natural products with a wide range of applications in cosmetics, food, medicine, agriculture, and biofuels. Heterologous biosynthesis is increasingly employed for sesquiterpene production, aiming to overcome the limitations associated with chemical synthesis and natural extraction. Sesquiterpene synthases (STSs) play a crucial role in the heterologous biosynthesis of sesquiterpene. Under the catalysis of STSs, over 300 skeletons are produced through various cyclization processes (C1-C10 closure, C1-C11 closure, C1-C6 closure, and C1-C7 closure), which are responsible for the diversity of sesquiterpenes. According to the cyclization types, we gave an overview of advances in understanding the mechanism of STSs cyclization from the aspects of protein crystal structures and site-directed mutagenesis. We also summarized the applications of engineering STSs in the heterologous biosynthesis of sesquiterpene. Finally, the bottlenecks and potential research directions related to the STSs cyclization mechanism and application of modified STSs were presented.

PosDiffNet: Positional Neural Diffusion for Point Cloud Registration in a Large Field of View with Perturbations

Point cloud registration is a crucial technique in 3D computer vision with a wide range of applications. However, this task can be challenging, particularly in large fields of view with dynamic objects, environmental noise, or other perturbations. To address this challenge, we propose a model called PosDiffNet. Our approach performs hierarchical registration based on window-level, patch-level, and point-level correspondence. We leverage a graph neural partial differential equation (PDE) based on Beltrami flow to obtain high-dimensional features and position embeddings for point clouds. We incorporate position embeddings into a Transformer module based on a neural ordinary differential equation (ODE) to efficiently represent patches within points. We employ the multi-level correspondence derived from the high feature similarity scores to facilitate alignment between point clouds. Subsequently, we use registration methods such as SVD-based algorithms to predict the transformation using corresponding point pairs. We evaluate PosDiffNet on several 3D point cloud datasets, verifying that it achieves state-of-the-art (SOTA) performance for point cloud registration in large fields of view with perturbations. The implementation code of experiments is available at https://github.com/AI-IT-AVs/PosDiffNet.

Multi-Modality Affinity Inference for Weakly Supervised 3D Semantic Segmentation

3D point cloud semantic segmentation has a wide range of applications. Recently, weakly supervised point cloud segmentation methods have been proposed, aiming to alleviate the expensive and laborious manual annotation process by leveraging scene-level labels. However, these methods have not effectively exploited the rich geometric information (such as shape and scale) and appearance information (such as color and texture) present in RGB-D scans. Furthermore, current approaches fail to fully leverage the point affinity that can be inferred from the feature extraction network, which is crucial for learning from weak scene-level labels. Additionally, previous work overlooks the detrimental effects of the long-tailed distribution of point cloud data in weakly supervised 3D semantic segmentation. To this end, this paper proposes a simple yet effective scene-level weakly supervised point cloud segmentation method with a newly introduced multi-modality point affinity inference module. The point affinity proposed in this paper is characterized by features from multiple modalities (e.g., point cloud and RGB), and is further refined by normalizing the classifier weights to alleviate the detrimental effects of long-tailed distribution without the need of the prior of category distribution. Extensive experiments on the ScanNet and S3DIS benchmarks verify the effectiveness of our proposed method, which outperforms the state-of-the-art by ~4% to ~ 6% mIoU. Codes are released at https://github.com/Sunny599/AAAI24-3DWSSG-MMA.

Fast-setting multi-stimuli responsive silk hydrogels tailored with dual network interface

Silk fibroin, a naturally occurring and versatile polymer with broad range of material forms and applications, presents numerous advantages. Notably, hydrogels are noteworthy among these advantages due to their inherent hydrophilicity, biocompatibility, and flexibility. Nevertheless, their practical utility has been hindered by intricate preparation procedures and limited mechanical strength. In this investigation, we introduce a straightforward and cost-effective method for producing a novel silk hydrogel. Subsequent to the hydrogel synthesis, an ethanol treatment was employed to induce the preferential formation of β-sheet conformations. The developed system exhibits rapid gelation (gelation time: 3 min) and non-swelling behavior (swelling ratio = 0.47). This suggests its potential suitability for biomedical devices without imposing strain on adjacent tissues. Conformational rearrangement of polymeric chains induced by ethanol treatment contributed to heightened material crystallinity, resulting in a remarkable 55.9% increase in Young's modulus and a substantial 91.2% enhancement in ultimate tensile strength for the ethanol-treated hydrogels (SF/AG/Gly/Eth) when compared to their non-ethanol-treated counterparts (SF/AG/Gly). Additionally, in-situ biodegradation studies on silk hydrogels confirmed the material's degradability, with a notable 94.75% weight loss within 3 days, following pseudo-second-order kinetics (R2 = 0.991). This degradation is attributed to action of proteolytic enzymes on the amino acid units of silk fibroin, alleviating concerns regarding waste generation. Moreover, these pH and water-responsive hydrogels can be harnessed in the fabrication of intelligent biomedical devices. Further, drug release studies on silk hydrogels showed that both ethanol-treated and non-treated samples displayed sustained drug release following the Korsmeyer−Peppas model, with an entrapment efficiency of 65.6% and 48.2%, respectively.

Dosimetry of Large Field Valencia applicators for Cobalt-60-based brachytherapy.

Non-melanoma skin cancer is one of the most common types of cancer and one of the main approaches is brachytherapy. For small lesions, the treatment of this cancer with brachytherapy can be done with two commercial applicators, one of these is the Large Field Valencia Applicators (LFVA). The aim of this study is to test the capabilities of the LFVA to use clinically 60Co sources instead of the 192Ir ones. This study was designed for the same dwell positions and weights for both sources. The Penelope Monte Carlo code was used to evaluate dose distribution in a water phantom when a 60Co source is considered. The LFVA design and the optimized dwell weights reported for the case of 192Ir are maintained with the only exception of the dwell weight of the central position, that was increased. 2D dose distributions, field flatness, symmetry and the leakage dose distribution around the applicator were calculated. When comparing the dose distributions of both sources, field flatness and symmetry remain unchanged. The only evident difference is an increase of the penumbra regions for all depths when using the 60Co source. Regarding leakage, the maximum dose within the air volume surrounding the applicator is in the order of 20% of the prescription dose for the 60Co source, but it decreases to less than 5% at about 1cm distance. Flatness and symmetry remains unaltered as compared with 192Ir sources, while an increase in leakage has been observed. This proves the feasibility of using the LFVA in a larger range of clinical applications.