Secondary organic aerosol (SOA) mass in the Southern USA during winter-spring 2022 were simulated by integrating Comprehensive Air quality Model with extensions (CAMx) with the UNIfied Partitioning-Aerosol phase Reaction (UNIPAR) model, which predicts SOA formation via multiphase reactions of hydrocarbons. UNIPAR streamlines multiphase partitioning of oxygenated products and their heterogeneous reactions by using explicitly predicted products originating from 10 aromatics, 3 biogenics, and linear/branched alkanes in different carbon lengths. UNIPAR simulations were compared with those using the partitioning-based model (SOAP), which uses simple surrogate products for each precursor. Both UNIPAR and SOAP showed similar tendencies in SOA mass but slightly underpredicted against observations at given five ground sites. However, SOA compositions and its sensitivity to environmental variables (sunlight, humidity, NOx, and SO2) were different between two models. In CAMx-UNIPAR, SOA was predominated by alkane, terpene, and isoprene, and was notably influenced by humidities showing high SOA concentrations with wet-inorganic salts, which accelerated aqueous reactions of reactive organic products. NO2 was positively correlated with biogenic SOA because elevated nitrate radicals effectively oxidized biogenic hydrocarbons at night and increased hygroscopic nitrate aerosol promoted SOA growth via organic heterogeneous chemistry. Anthropogenic SOA, which formed mainly via the daytime oxidation with OH radicals, was weakly and negatively correlated with NO2 in cities. In CAMx-UNIPAR, the sensitivity of SOA to aerosol acidity (neutral vs. acidic aerosol at cation/anion = 0.62) was dominated by isoprene SOA. The decline of NOx emission benefits the mitigation of SOA burdens in the Southern USA where biogenic hydrocarbons are abundant.

"Advanced Math for Physics" course - All Equations  v.2024-01-17  SummaryThis Authorea document, written by Michiel Schotten (Dolphinity, the Netherlands), contains all the mathematical equations in LaTeX format from the "Advanced Math for Physics" course created by Ville Hirvonen from "Profound Physics" at https://profoundphysicscourses.com/courses/vector-calculus-for-physics-a-complete-self-study-course-copy/ (before a major update of this course was done in January 2024).  The purpose of this document is to create downloadable PDFs of the online course material, a project commissioned by Profound Physics and carried out by Dolphinity.Background & MethodThe reason Authorea was chosen for this project is because it allows to import the equations from the course material in LaTeX format, which are beautifully rendered as the correct equations in Authorea; and then to export it into MS-Word, where the equations are automatically converted into MS Word's Equation fields (a unique feature that Authorea provides).  Each MS-Word equation can then be inserted into the appropriate place into the copied course material, with customizable layout and formatting in MS-Word, and then saved as a separate PDF document for each lesson.The steps taken to create this Authorea document are:Copy/paste the HTML of each lesson from the course website into MS-WordUsing an Advanced Find function in MS-Word, find the HTML code of each equation, and copy/paste that piece of HTML code into MS-Excel (so one equation per MS-Excel row, copy/pasting them one by one)Using customized MS-Excel formulas, convert the HTML code of each equation (including the URL-specific character codes that it contains) into the proper LaTeX format(For the URL character conversion, I used https://www.urldecoder.org/ )Using the "Concatenate" formula in MS-Excel, wrap the LaTeX code of each equation within "\[...\]\\" characters, so that each equation will be rendered in Authorea in the "Display" format and with one line of space in between each equationIn the Authorea document, from the top menu select:  Insert/LaTeXWithin the newly created block of LaTeX in Authorea, copy/paste the desired block of MS-Excel cells with the wrapped LaTeX equationsClicking outside of the LaTeX block will have Authorea properly render these equationsExport this Authorea document to MS-Word (where the LaTeX equations will be automatically converted to MS-Word Equation fields), using "Export" in the top right cornerFor additional information, see this excellent tutorial on how to use LaTeX within Authorea: https://www.authorea.com/users/77723/articles/110898-how-to-write-mathematical-equations-expressions-and-symbols-with-latex-a-cheatsheet .

As oceanic moisture evaporates, it leaves a signature on sea surface salinity. Roughly 10% of the moisture that evaporates over the ocean is transported over land, allowing the salinity fields to be a predictor of terrestrial precipitation. This research is among the first in published literature to assess the role of sea surface salinity for improved predictions on low-skill summertime subseasonal timescales for terrestrial precipitation predictions. Neural networks are trained with the CESM2 Large Ensemble using North Atlantic salinity anomalies to quantify predictability of U.S. Midwest summertime heavy rainfall events at 0 to 56-day leads. Using explainable artificial intelligence, salinity anomalies in the Caribbean Sea and Gulf of Mexico are found to provide skill for subseasonal forecasts of opportunity, e.g. confident and correct predictions. Further, a moisture-tracking algorithm applied to reanalysis data demonstrates that the regions of evaporation identified by neural networks directly provide moisture that precipitates in the Midwest.

The Zhurong rover, which played a crucial role in China’s Mars exploration mission, accomplished a successful landing on the southern surface of the Martian Utopia Basin on May 15, 2021. The ground-penetrating radar(GPR) carried by the Zhurong rover has a dual channel design, allowing it to effectively detect the morphology and geological structure of the subsurface in the landing area. The low-frequency data reveals a multi-layered underground structure that stretches 80 meters beneath the surface along the path of the Mars rover. The high-frequency data reveals the reshaping of Martian surface sediments caused by meteorite impacts and weathering in the soil layer up to a depth of 5 meters underground. Beneath the weathering layer, a shallow stratigraphic sequence is formed by the overlapping of sedimentary rock units. The findings from the research using GPR data serve as a crucial foundation for comprehending the geological history of the Martian Utopia Basin.

Relative localization is an essential part of autonomous multi-agent systems. Existing methods often require real-time communications, pre-installed infrastructure, and substantial computational resources. In this study, drawing inspiration from the collective behaviors of primitive animals, we propose an infrastructure-free 2D distributed relative localization framework utilizing onboard ranging sensors. We start with system modeling, based on which optimal sensor configuration and algorithm design are conducted. Subsequently, we perform a thorough performance analysis and validate the overall system design through field tests using unmanned ground vehicles (UGVs) equipped with ultra-wideband (UWB) ranging sensors and micro-controller units onboard. Contributions include the following: the geometric dilution of precision (GDOP) and Cramér-Rao lower bound (CRLB) are derived; a novel Euclidean distance matrix (EDM)-based trilateration algorithm and a maximum likelihood estimation algorithm are proposed; and comprehensive simulation and field tests are conducted to validate the viability of the proposed framework. Two use cases are considered: to localize a target sensor and to localize an agent. The theoretical, numerical, and experimental results will shed light on the design and optimization of relative localization systems, and our proposed framework holds potential for future extensions to 3D scenarios, different unmanned vehicle platforms, and multi-robot cooperative systems.

Surge glaciers have a unique type of glacial acceleration, surging, in which the glacial system leaves a period of quiescence and experiences velocities that are up to 200 times the non-surge velocities. Surge events play a critical role in sea level rise (SLR), as the mass loss from even a single marine-terminating glacier during a surge has been estimated to be upwards of 0.5 percent of annual global SLR.Glacial hydrology, the water that flows through and below the glacier, plays a critical role in surge evolution and initiation, as the initiation of a surge requires decoupling of the glacier from the bed via reduction of basal friction, which is directly related to the hydropotential and water accumulation at the basal boundary.This work establishes a simple framework for accurately determining the basal pre-surge hydrologic conditions and modeling the subsequent glacial dynamics, which may lead to a surge event.This work takes a combined approach via image classification, algorithmic interpretation of ICESat-2 altimetry data, and 3D glacial modeling. Using satellite imagery, the  distribution of surface water is determined via a simple image classification approach.Distribution of glacial surface water is determined via supervised classification of satellite imagery. The volume of surface water is determined by estimating water and ice surface elevation for each water feature with the Density-Dimension Algorithm for ice surfaces. The DDA-ice-2 determines ice surface height, crevasse morphology of wet and dry crevasses and water depth from ICESat-2 ATLAS data. The DDA-bifurcate algorithm determines ice surface height, melt pond morphology, and water depth from ICESat-2 ATLAS data.

Globally, a considerable number of infants and children younger than five are falling victim to diarrheal diseases predominantly caused by rotaviruses, which are non-enveloped, double-stranded RNA (dsRNA) viruses able to cause acute gastroeteritis and extragastrointestinal complications. Annually, human rotaviruses cause two million hospitalizations and over 500,000 deaths worldwide. Rotaviral replication, pathogenesis, and immune evasion are propagated by non-structural protein 1 (NSP1), encoded by segment five of their dsRNA genome. We examined 60 urine and stool samples from children aged 2-60 months admitted to an Obstetric and Children Hospital in Babylon over a 60-day time period with the diagnosis of acute group A rotavirus gastroenteritis. This study aimed to check the presence of NSP1 by immunochromatography assay and RT-qPCR. Immunochromatography assay detected NSP1 in 100% of urine and stool samples; however, RT-qPCR only detected it in 66.7% of urine and 50% of stool samples. RT-qPCR found 12 out of 30 urine and stool samples positive, accounting for 40% of participants. No significant correlations between RT-qPCR results and sociodemographic factors were found. Results found 73.3% of acute gastroenteritis cases were in children under two. Additionally, the urinary detection of NSP1 suggests that rotaviruses may cause extra-gastrointestinal infections, e.g., systemic infection or viremia.

Escherichia coli (E. coli) strains have been classified into eight distinct phylogenetic clusters as per a novel quadruplex PCR method. Nevertheless, the precise phylogenetic relationship among these bacterial lineages remains unclear. The Clermont phylotyping method has been utilized by the current study to further clarify E. coli phylogenetic clusters and assess resistance to antibiotics showed by uropathogenic E. coli (UPEC) strains in Iraq. Forty-two UPEC isolates were assessed for antibiotic sensitivity through a disk diffusion test, and the novel Clermont phylotyping method was utilized for the phylogenetic identification of isolates. The research findings revealed the varying prevalence of distinct phylogroups at the hospitals of Babylon province, Iraq, with Phylogroup B2, as the predominant group accounting for 47.61%, followed by Clade I (14.28%), B1 (11.90%), A (9.52%), D (4.76%), C (2.38%), and an unidentified phylogroup (9.52%). Additionally, out of the 42 Uropathogenic E. coli isolates studied, 37 (88.09%) showed multidrug resistance (MDR) and 5 isolates (11.90%) displayed extensive drug resistance (XDR). MDR and XDR strains within Phylogroup B2 accounted for 17 out of 37 cases (45.24%) and 3 out of a total of 5 cases (60%), respectively, indicating a large proportion of MDR and XDR UPEC isolates within phylogroup B2. Moreover, two new phylogroups, namely C and clade I, were identified, linked respectively to E. coli sensu stricto and cryptic E. coli. Thus, further studies are required to be conducted elsewhere to gain a better perception of both antibiotic-resistance characteristics and the occurrence of diverse phylogroups in Iraq.

Automatic Speech Recognition (ASR) is an area of research that's constantly evolving, thanks to important advancements like machine learning and deep learning techniques. Its applications are wide-ranging, touching fields like healthcare, public services, and interfaces between humans and machines. What's particularly noteworthy is the pressing need for highquality Arabic datasets to enhance the capabilities of speech recognition on devices that use the Arabic language. In this paper, we introduce a new dataset created with great care, designed specifically for recognizing Arabic speech when spoken by Tamazight speakers. This effort significantly broadens the pool of linguistic resources available for research and practical use. A crucial aspect of developing this dataset is the rigorous quality control applied to the data, which, in turn, improves the accuracy and effectiveness of Arabic speech recognition models. By making use of this innovative dataset, we enable the creation and evaluation of Arabic ASR systems tailored precisely to the needs of Tamazight speakers. This addresses a critical gap in the field of Arabic speech recognition, as it focuses on linguistic groups that have been underrepresented in this technology.

A document by Hamid Mottaghi Golshan. Click on the document to view its contents.

Land-atmosphere feedbacks act through process chains that link variables in the land-atmosphere system. For the global energy and water cycles, the first link in the chain is soil moisture. Flux tower sites provide in-situ observations, including land surface states, surface fluxes, and nearsurface atmospheric states, to validate these links; however, they are unevenly distributed over the globe. Therefore, to obtain a global view of observationally based land-atmosphere coupling metrics, satellite data are useful. Among satellite products, the Soil Moisture Active Passive (SMAP) satellite provides the closest match to in-situ observations. However, SMAP exhibits stochastic random noise that can deflate coupling estimates. Since soil moisture variability closely follows a first-order Markov process, it typically has a distinct red noise spectrum. Satellite data with random noise has a whiter spectrum at high frequencies that can be compared to the expected red spectrum. Also, missing data in SMAP are not entirely random; its 8-day repeating polar orbit creates a cadence of missing data for both ascending and descending overpasses, depending on the location. This creates additional artifacts in the power spectrum, calculated through lagged autocovariance in the time series, with harmonic spikes at 8, 4 (8/2), 2 2/3 (8/3), and 2 (8/4) days that broaden due to the satellite's orbital variations. To be optimally useful for quantifying land-atmosphere feedbacks, the effects of random noise and periodic missing data must be minimized. A power spectrum adjustment technique has been designed to remove the orbital harmonic spikes from Level 3 (L3) SMAP data. This is achieved by fitting and removing a catenary function to the power spectrum between harmonic spikes. This adjusted spectrum is then scaled to match surface layer soil moisture observations at sites of the AmeriFlux network (in-situ data), which exhibit relatively low noise and have spectra that are very similar to those produced by offline land surface models (LSMs). Utilizing validated spectral data from gridded LSM-based datasets, a global L3 SMAP product with removed noise and harmonic effects is being produced. We will present results quantifying the extent to which this technique improves SMAP data and its temporal correlation with observations.

A document by Palak Handa. Click on the document to view its contents.

The idea of data warehousing is data model independence and does not require the adoption of any data model. Nevertheless, relational databases are assumed to be used in the common notion of data warehouses. Typically, it is thought that data warehouses are relational data storage, and relational database management systems (RDBMSs) are used to process them. Data warehousing frequently features Structured Query Language (SQL) reporting capabilities, enabling access to the data in a standard manner. "Not only SQL" (NoSQL) databases are viewed as high-speed data structures appropriate for filtering and lookup operations suitable for complex processing tasks. Traditional database systems are incapable of dealing with vast volumes of data for knowledge discovery and complex analytics. This incapability is facing a paradigm shift in technologies, techniques, concepts, and methods. The key problem is to achieve a good balance between the characteristics of classical data warehouses employing relational database management systems and the potential afforded by NoSQL database management systems in a big data environment. This paper covers the integration of disparate big data technologies using an opensource NoSQL columnar database management system to address the possibility of constructing data warehouse (DW) solutions in a big data environment.

A document by A S Bhattacharyya. Click on the document to view its contents.

Virtual memory compliance refers to the extent to which the behaviour of the virtual memory subsystem adheres to the specifications outlined in the RISC-V architecture. It ensures that the behavior of virtual memory operations, particularly the manipulation of permission bits within page table entries, aligns accurately with the documented expectations and standards of the RISC-V architecture. We developed customized RISC-V assembly tests that systematically modify permission bits within page table entries. These tests are compiled using GNU toolchain and executed using the RISC-V assembly code on both the Spike and Sail simulators through the RISCOF framework, facilitating cross-platform analysis. By comparing the resulting log files from the two simulators, we discern any inconsistencies or variations in memory access behaviors. The findings of this investigation provide insights into the fidelity of RISC-V architecture specifications with respect to virtual memory operations. We generally worked with two level page table in which first level is named as level 1 and second level is named as level 0. These outcomes are synthesized into a comprehensive HTML report, offering an in-depth exploration of permission bit effects on virtual memory within varying privilege modes. This research enhances comprehension of virtual memory functionality in the RISC-V architecture. This study contributes to a more robust understanding of virtual memory behavior and engenders confidence in the modeling and simulation of such systems.

Effective crack detection is pivotal for the structural health monitoring and inspection of buildings. This task presents a formidable challenge to computer vision techniques due to the inherently subtle nature of cracks, which often exhibit low-level features that can be easily confounded with background textures, foreign objects, or irregularities in construction. Furthermore, the presence of issues like non-uniform lighting and construction irregularities poses significant hurdles for autonomous crack detection during building inspection and monitoring. Convolutional neural networks (CNNs) have emerged as a promising framework for crack detection, offering high levels of accuracy and precision. Additionally, the ability to adapt pretrained networks through transfer learning provides a valuable tool for users, eliminating the need for an in-depth understanding of algorithm intricacies. Nevertheless, it is imperative to acknowledge the limitations and considerations when deploying CNNs, particularly in contexts where the outcomes carry immense significance, such as crack detection in buildings. In this paper, our approach to surface crack detection involves the utilization of various deep learning models. Specifically, we employ fine-tuning techniques on pre-trained deep learning architectures: VGG19, ResNet50, Inception V3, and EfficientNetV2. These models are chosen for their established performance and versatility in image analysis tasks. We compare deep learning models using precision, recall, and F1 score.

Most of the drift velocity models that exist have limitations, as they were developed based on experiments that hardly considered the conjunction of liquid properties and pipe geometry effects. Additionally, some of the models are formed complexly and several parameters require to be optimized. This study focuses on the application of the artificial neural network (ANN) model in predicting the drift velocity of Taylor bubbles rising in stagnant fluids through horizontal and inclined pipes. A comprehensive experimental database including 364 data points from the open literature was applied to develop the ANN model. Inclination angle, pipe diameter, liquid density, liquid viscosity, and surface tension were used as input variables. Finally, a multilayer perceptron (MLP) feed-forward backpropagation neural network having 12 neurons in the hidden layer was selected as the optimized ANN model that showed the best performance for the prediction of drift velocity. The obtained ANN model showed superior performance in comparison with the support vector machine (SVM) model and four drift velocity correlations, with high accuracy for the training data set (MSE=0.000127, R2=0.9985, and MAPE= 5.85%), testing data set (MSE=0.00028, R2=0.9983, and MAPE= 5.45%), and all data (MSE=0.000137, R2=0.9982, and MAPE= 5.69%).

This study aims to identify and demonstrate the steps involved in the accounting treatment of the tax consolidation technique for value-added tax in a corporate group in accordance with the financial accounting system. This is achieved using both descriptive and analytical approaches to describe the study variables, read and analyse data, and interpret legislative texts. As for the field aspect of the study, a case study of the Condor group was adopted. The study reached several results, the most important of which are: allocating a dedicated account for the group's operations in the financial accounting system; and that the tax consolidation of the value-added tax helps to avoid financial hardship and provide consistent liquidity for the group's companies. The study also recommended, in the end, simplifying the conditions for forming tax groups and increasing interest in group accounting in the academic and professional environment.

Marine forecasts are essential for safe navigation, efficient offshore operations, coastal management, and research, especially in areas with a such harsh conditions as the Arctic Ocean. They require accurate predictions of ocean currents, wind-driven waves, and other oceanic parameters. However, physics-based numerical models, while precise, are computationally demanding. Consequently, data-driven methods, which are less resource-intensive, may offer a more efficient solution for sea state forecasting. This paper presents an analysis and comparison of three data-driven models: our newly developed convLSTM-based MariNet, FourCastNet and the PhydNet, a physics-informed model for video prediction. Using metrics such as RMSE, Bias and Correlation, we demonstrate the areas where our model surpasses the performance of the prominent prediction models. Our model achieves improved accuracy in forecasting ocean dynamics compared to FourCastNet and PhyDNet. We also find that our model requires significantly less training data, computing power, and consequently provides less carbon emmisions. The results suggest that data-driven models should be further explored as a complement to physics-based models for operational marine forecasting. They have the potential to enhance prediction accuracy and efficiency, enabling more responsive and cost-effective forecasting systems.

The Spectre of Generative AI Over Advertising, Marketing, and BrandingBibhuti Bhusan RoutrayAbstractThe rapid progress of generative artificial intelligence (AI) technologies such as DALL-E, GPT-3, and ChatGPT has sparked discussions about the potential impacts on creative industries. With their ability to synthesize novel text, images, music, and other content, these AI systems could automate certain creative tasks. Some predict this will disrupt creative professions like advertising and branding. This paper provides an in-depth review of academic literature and industry commentary on the capabilities and limitations of current generative AI systems. It analyzes the potential near-term impacts on the advertising and branding fields, where the ability to generate marketing content and assets could significantly disrupt workflows and employment.However, the paper argues generative AI is unlikely to wholly replace human creativity and strategic judgment in these industries in the foreseeable future, given limitations in areas like intentionality, consistency, bias, and assessment of aesthetic qualities. While automation of some discrete tasks is probable, generative AI currently lacks the holistic creative abilities that define human ingenuity. As such, the paper suggests the medium-term impacts may be concentrated in partial automation of technical execution and the need for adaptation by creatives.Realizing the benefits of AI augmentation while mitigating harms will require responsible governance and development of these transformative but still limited technologies. With prudent policies and efforts to develop complementary human skills, generative AI can positively assist advertising, branding, and creativity at large. But sustaining the uniquely human gifts of strategy and aesthetic judgment necessitates building AI as an ally to creatives rather than adversary. If fostered thoughtfully, this current wave of innovation can propel human imagination and productivity to new heights.IntroductionGenerative artificial intelligence (AI) has rapidly advanced in capabilities in recent years. Systems like OpenAI’s image generator DALL-E (Ramesh et al., 2022), text generator GPT-3 (Brown et al., 2020), and conversational agent ChatGPT (Thoppilan et al., 2022) point to a future where AI can synthesize novel content like images, text, video, and more. The creative potential demonstrated by these systems has led to both excitement about new possibilities for augmenting human creativity, and also concerns about potential disruption of creative industries (Sevilla and Vishwanath, 2022).Fields like advertising and branding are particularly implicated, given their reliance on crafting novel content and messaging. Commentators warn creatives in these fields face being automated out of jobs by AI that can generate ads, logos, and brand content (Chavez, 2022; Kulkarni, 2022). However, research suggests generative AI’s actual capabilities and limitations indicate it will not wholly replace human creatives in the near future. This paper reviews literature on the abilities and constraints of current generative AI systems, and their likely impacts on advertising and branding. It argues that while generative AI does present challenges for these creative fields, with prudent regulation and responsible development, it offers opportunities to complement human creativity more than replace it.Generative AI’s CapabilitiesRecent advances in generative AI, especially in natural language processing (NLP) and computer vision, have dramatically expanded what these systems can autonomously produce. Models like DALL-E 2 and Imagen can generate photorealistic images from text prompts (Ramesh et al., 2022; Saharia et al., 2022). GPT-3 exhibits an unprecedented ability to produce human-like text for a wide variety of writing tasks (Brown et al., 2020). AI chatbots like Google’s LaMDA and Anthropic’s Claude can hold natural-language conversations (Thoppilan et al., 2022).These capacities have enabled new creative applications. DALL-E 2 has been used by artists to compose novel images as inspiration for artworks (Soo, 2022). GPT-3 can generate marketing copy, slogans, logos, and other advertising and branding content from basic prompts (Subramanian, 2022). Chandrasekaran et al. (2022) demonstrate AI’s potential for automating graphic design tasks like generating logos, posters, and brochure layouts. Music generators like Anthropic’s Claude Aria compose original songs and instrumentals (Lynch, 2022).The speed and scale at which these systems can produce novel, often high-quality output suggests the beginnings of an AI revolution in creative industries. Why pay a copywriter when an AI can generate hundreds of slogans in seconds? Why hire a designer when an AI can instantly generate logos and other branding material? Commentators argue generative AI can replicate tasks across advertising, marketing, design, music, architecture, and other creative fields, threatening the livelihoods of human creatives (Sample, 2022; Sevilla and Vishwanath, 2022).Limitations of Current Generative AIHowever, examinations of current generative AI systems reveal meaningful limitations that likely preclude full automation of creative work in the near future. While their capabilities are impressive, these technologies do not yet match human creativity and intuition.One limitation is a lack of intentionality. Though they can produce novel content, current AI systems do not truly understand the meaning and context of what they generate (Bommasani et al., 2021). This makes it challenging for them to consistently create content for a specific purpose or goal. In branding, advertisers want content that strategically builds the brand; AI cannot yet replicate this intentionality (Subramanian, 2022).Generative AI also struggles with consistency and coherence in its output. Minor prompt variations can yield unpredictable or nonsensical results. DALL-E 2 images sometimes contain visual flaws and artifacts revealing their artificial origins (Ramesh et al., 2022). GPT-3 output tends to lose coherence over long passages, reflecting its statistical, probabilistic approach to text generation (Adiwardana et al., 2020). These inconsistencies reduce generative AI’s current utility for tasks demanding high reliability.Data biases also constrain current systems. Models like DALL-E and GPT-3 are trained on vast datasets scraped from the Internet, inheriting human biases around things like race, gender, and ethnicity embedded in that data (Bommasani and Cardie, 2021). Efforts to reduce biases are ongoing, but remain a challenge. Generated content sometimes reflects and amplifies these biases in problematic ways (Abid et al., 2021).Finally, major limitations exist in evaluating and critiquing generative AI output. Systems lack mechanisms for assessing aesthetic qualities like creativity or originality. GPT-3 cannot critically judge its own text; it relies on human feedback. Current systems also cannot explain or provide rationale for their output. The ”black box” nature of generative AI output makes it hard to improve (Song and Shu, 2022).These limitations highlight that while generative AI can replicate certain discrete creative tasks, truly matching human creativity and intuition remains challenging. As Cross (2022) argues, current systems lack the intentionality, appreciation of aesthetics, critical thinking, and cultural awareness true creativity demands. Overcoming these limitations to accurately emulate human creativity likely requires fundamental advances in artificial general intelligence.Though limitations exist, some brands have already begun experimentally using generative AI in marketing and advertising campaigns. Documenting use cases helps illuminate current capabilities and challenges.Examples include AI-generated social media ads for the non-profit Aspiration (Del Ray, 2022). The tool Anthropic Claude has been used to generate product descriptions for Shopify merchants (Subramanian, 2022). Agencies have tested AI-composed background music for ad spots, with mixed aesthetic results (Jain, 2021). Generative copywriting apps aim to help marketers produce more content faster, though quality control remains a concern (Aron, 2022).Early experiments highlight issues around branding coherence and consistency, as different prompts yield unpredictable output. But they also demonstrate utility for rapid content prototyping. Case studies suggest a hybrid approach combining AI-generated raw material and human creative refinement offers the most promise currently. More research is needed on actual marketing outcomes using generative AI.Impacts on Advertising and BrandingGiven its limitations, current generative AI on its own is unlikely to wholly displace human creatives in advertising and branding. However, it does present meaningful impacts for these fields that necessitate adaptation.One likely effect is partial automation of specific creative tasks. Generative AI shows aptitude for rote work like generating raw copy and imagery that human creatives can then refine (Subramanian, 2022). Systems like DALL-E 2 and GPT-3 may increasingly act as ”digital assistants” to boost individual productivity (Sevilla and Vishwanath, 2022).This may displace some entry-level creative roles focused on basic content generation. The economic impacts may be uneven, as middle and top-tier creatives who focus more on strategy, leadership, and relationships could remain in demand even with automation of some tasks (Thompson, 2022). Those relying solely on technical execution face more risk.Generative AI also presents new medium-term challenges around branding, intellectual property, and liability. Widely available text and image generators make plagiarism and unauthorized use of brand assets easier (Chavez, 2022). Creators already struggle to assert copyright given AI’s ability to remix and synthesize new content (Sample, 2022). Attributing ”authorship” and responsibility for AI output is also legally murky, presenting challenges if inappropriate or harmful content is created (Hristov, 2021).At the same time, used ethically and responsibly, generative AI could benefit branding and advertising. Assistants like GPT-3 allow faster iteration of ideas and content (Subramanian, 2022). Democratized access to design tools could help small brands afford quality aesthetics (Chandrasekaran et al., 2022). Responsibly deployed systems could also monitor for harmful biases and promote diversity in generated content. Realizing these benefits likely requires proactive self-regulation by developers and prudent government oversight (Leslie, 2019).Given its disruptive potential, advertising and branding professionals must proactively adapt their skills and working methods to integrate generative AI effectively. This necessitates developing AI literacy to capitalize on strengths while mitigating limitations (Lewis, 2022). Rather than resist change, creatives should explorer collaborating with AI tools on defined tasks.Agencies and professional associations can assist with training programs on incorporating AI responsibly into creative workflows. Developing hybrid roles at the intersection of creative direction and technical AI proficiency can help retain unique human skills while staying competitive (Thompson, 2022). AI should augment individual productivity and ideation rather than replace creative roles.For displaced entry-level professionals, support and retraining will be critical. Building transferable skills in strategy, ideation, and project management can open adjacent career pathways. Lifelong learning mindsets help creatives stay adaptable even amidst major technological shifts. Investment from tech firms and governments in just workforce transitions can mitigate negative impacts on displaced creatives.The Path ForwardIn assessing generative AI’s impacts on creativity, Cross (2022, p.120) advises avoiding both ”unjustified exuberance” and ”excessive dystopian fears.” Current systems are impressive but limited; they do not yet truly replicate human creativity. But they present meaningful challenges for advertising, branding, and other creative fields that necessitate adaptation.Rather than full automation, the medium-term future likely involves AI and humans complementing one another - with creatives using AI tools to augment their workflows rather than replace them. But to maximize the benefits and mitigate risks, responsible governance of these technologies is required, promoting innovation while managing disruptions and downsides. Education to develop AI literacy in both creatives and consumers can also help intelligently integrate these technologies.Continued progress in coming years is likely as computational power grows exponentially. But research suggests truly human-like creative intelligence remains a distant prospect. With prudent regulation and responsible development, generative AI can positively assist human creativity more than destroy creative professions - presenting opportunities to imagine new ways of blending human ingenuity with machine ability.Realizing the promise of AI to augment creativity requires responsible design and use of these technologies. Developers should proactively assess risks and benefits, engage affected communities, and enact controls to address issues like biases in training data (Leslie, 2019). Generative models should be transparent in capabilities and limitations to prevent overreach or deception.Brands must also use AI ethically, considering consumer impacts and creative rights. Generative content should be clearly identified as such, rather than passed off as original human work. Strategic usage that respects creative professionals is important for sustainable integration. Codes of conduct, audits, and oversight bodies can help scale responsible practices (Bommasani et al. 2021). With conscientious governance and norms, generative AI can positively transform creative industries.ConclusionThe emergence of powerful generative AI systems represents a potentially disruptive technological change for creative fields like advertising and branding. Technologies like DALL-E, GPT-3, and Claude display unprecedented ability to autonomously synthesize novel text, images, music, and more. However, a sober analysis reveals current limitations in matching human creativity and intentionality.While these technologies can automate certain discrete creative tasks, they lack the aesthetic sensibility, cultural awareness, and strategic purpose that define human ingenuity. As such, in the near future generative AI is unlikely to completely displace professional creatives. The bigger near-term impacts will be partial automation of rote creative work and the need for adaptation by advertising and branding to integrate these technologies responsibly.Realizing the benefits of AI-enabled augmentation of creativity, while mitigating the risks, will require judicious governance and development of these rapidly advancing technologies. Artists, developers, policymakers, brands, and the public must collaborate to shape norms and policies that foster innovation but curb harmful misuse. Investing in educational initiatives to develop AI literacy and human-centered skills can equip the next generation of creatives to thrive alongside AI tools.With prudent and ethical deployment, generative AI can positively transform creative sectors by reducing drudgery and accelerating human imagination and productivity. But sustaining the uniquely human gifts of strategy, meaning, and aesthetic judgment in advertising and branding necessitates building these technologies as allies rather than adversaries to human creativity. If society approaches generative AI thoughtfully, this innovation wave can propel creativity to new heights by leveraging the complementary strengths of human and machine intelligence.ReferencesAbid, A., Farooqi, M., Zou, J., 2021. Persistent Anti-Muslim Bias in Large Language Models. arXiv preprint arXiv:2101.05783.Adiwardana, D., Luong, M.T., So, D.R., Hall, J., Fiedel, N., Thoppilan, R., Yang, Z., Kulshreshtha, A., Nemade, G., Lu, Y., Le, Q.V., 2020. Towards a human-like open-domain chatbot. arXiv preprint arXiv:2001.09977.Bommasani, R., Cardie, C. 2021. Assessing the State of the Art in AI Ethics and Governance. In: Larson, K., Winfield, A., Iphofen, R. (Eds.) The Matter of Trust: Ethical AI for Public Good. Springer, Cham, pp. 143–160.Bommasani, R., Hudson, D.A., Adeli, E., Altman, R., Arora, S., von Arx, S., Bernstein, M.S., Bohg, J., Bosselut, A., Brunskill, E., Brynjolfsson, E., Buch, S., Card, D., Castellon, R., Chatterji, N.S., Chen, A., Creel, K., Davis, J.Q., Demszky, D. and Dragan, A., 2021. On the opportunities and risks of foundation models. arXiv preprint arXiv:2108.07258.Brown, T.B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A. and Agarwal, S., 2020. Language models are few-shot learners. Advances in neural information processing systems, 33, pp.1877-1901.Chandrasekaran, A., Yadav, D., Chattopadhyay, A., Prabhu, V., Prabhakaran, B., 2022. GraphicAI: A platform for automated graphic design. Proceedings of the ACM Conference on Human Factors in Computing Systems (CHI). https://doi.org/10.1145/3491102.3517586Chavez, C., 2022. AI will kill the advertising creatives. Forbes, August 2. https://www.forbes.com/sites/chrischavez/2022/08/02/ai-will-kill-the-advertising-creatives/Cross, N., 2022. Has AI toppled the creatives? Design Studies, 76, pp.101-126.Hristov, K., 2021. Artificial intelligence and the copyright dilemma. IPWatchdog. https://www.ipwatchdog.com/2021/05/23/artificial-intelligence-copyright-dilemma/id=133389/Kulkarni, A., 2022. How AI is disrupting the creative industry. Forbes, July 5. https://www.forbes.com/sites/forbestechcouncil/2022/07/05/how-ai-is-disrupting-the-creative-industry/Leslie, D., 2019. Understanding artificial intelligence ethics and safety: A guide for the responsible design and implementation of AI systems in the public sector. The Alan Turing Institute.Lynch, J., 2022. Anthropic develops AI that can generate music. VentureBeat, November 16. https://venturebeat.com/ai/anthropic-develops-ai-that-can-generate-music/Ramesh, A., Pavlov, M., Goh, G., Gray, S., Voss, C., Radford, A., Chen, M., Sutskever, I., 2022. Zero-shot text-to-image generation. arXiv preprint arXiv:2102.12092.Saharia, C., Chan, W., Saxena, S., Li, L., Whang, J., Denton, E., Ghasemipour, S.K., Ayan, B., Balaji, S., Lopes, J.G.O. and Maharaj, T., 2022. Photorealistic text-to-image diffusion models with deep language understanding. arXiv preprint arXiv:2205.11487.Sample, I., 2022. AI art has a creativity problem even if it can mimic artists’ styles. The Guardian, August 29. https://www.theguardian.com/technology/2022/aug/29/ai-art-creativity-problem-mimic-artists-dalle-2Sevilla, A.J. and Vishwanath, A., 2022. The Creative Destruction Brought About by Generative AI. Harvard Business Review, 12.Soo, Z., 2022. How Artists Are Using AI Tools to Generate Images—and Questions. Artsy, September 7. https://www.artsy.net/article/artsy-editorial-artists-ai-tools-generate-imagesand-questionsSong, Y., Shu, R., 2022. Generative AI: What makes human-AI collaboration thrive? XRDS: Crossroads, The ACM Magazine for Students, 29(2), pp.46–51.Subramanian, S., 2022. How creatives and marketers feel about AI content creation. Think with Google, March. https://www.thinkwithgoogle.com/consumer-insights/ai-content-creation-creatives-marketers-sentiment/Thompson, D., 2022. What happens to creativity in an age of automation? The Atlantic, November 21. https://www.theatlantic.com/ideas/archive/2022/11/ai-artificial-intelligence-automation-work-creativity/672254/Thoppilan, R., De Freitas, D., Hall, J., Sharma, A., Felt, H., Nemade, G., Lu, E., 2022. LaMDA: Language Models for Dialog Applications. arXiv preprint arXiv:2201.08239.

Space-based measurements of the Earth's magnetic field with a good spatiotemporal coverage are needed to understand the complex system of our surrounding geomagnetic field. High-precision magnetic field satellite missions form the backbone for sophisticated research, but they are limited in their coverage. Many satellites carry so-called platform magnetometers that are part of their attitude and orbit control systems. These can be re-calibrated by considering different behaviors of the satellite system, hence reducing their relatively high initial noise originating from their rough calibration. These platform magnetometer data obtained from non-dedicated satellite missions complement the high-precision data by additional coverage in space, time, and magnetic local times. In this work, we present an extension to our previous Machine Learning approach for the automatic in-situ calibration of platform magnetometers. We introduce a new physics-informed layer incorporating the Biot-Savart formula for dipoles that can efficiently correct artificial disturbances due to electric current-induced magnetic fields evoked by the satellite itself. We demonstrate how magnetic dipoles can be co-estimated in a neural network for the calibration of platform magnetometers and thus enhance the Machine Learning-based approach to follow known physical principles. Here we describe the derivation and assessment of re-calibrated datasets for two satellite missions, GOCE and GRACE-FO, which are made publicly available. We achieved a mean residual of about 7 nT and 4 nT for low- and mid-latitudes, respectively.

An AI based Online Proctoring System isn't a new concept and many such capable exam portals do already exist. However, all of them have an unsolved design flaw which is server side processing.To detect any suspicious activity, the sites either take a snapshot of the examinee in regular intervals which is doable but is very weak, or they continuously send the video feed over to the server for processing which being comparatively more effective, is highly expensive. Sending video feeds of tens of thousands of students and processing them in real time can be very heavy on the server as well as costly for the client. To counter all these flaws, proposing an AI based proctoring system that securely works on the client side. Overall goal is to allow the face detection system and suspicious activity detection system to run on the client side which will significantly reduce the server load and dependency on the network. In this review paper we explored various algorithms for face verification, object detection, also reviewed pre-existing OPS systems and learned about their architecture.

Dissolved strontium (Sr) concentrations in the Southern Ocean water samples and Sr export fluxes from sediment trap moorings at 1000 m were used to assess particulate organic carbon (POC) export associated with Acantharia for 2010, 2018 and 2020. The dissolved Sr data revealed a prominent vertical gradient with lower surface Sr concentrations depleted up to 1.4% relative to deep waters. A strong latitudinal surface gradient was observed, ranging from 87.3 near the northern end to 88.5 near the southern end of a transect through the Australian sector of the Southern Ocean. These findings highlight the significant role that Acantharia, which precipitate celestite (SrSO4), play in marine Sr cycling. Seasonal variability in Sr export fluxes can be large, particularly during intense events in summer, and reaches a maximum of 2.8 , contributing up to 7% of the POC export flux. The coincidence of Sr flux with the second peak of POC export flux implies a potential association of Acantharia biomass with summertime productivity.

In this paper, we propose a novel transmission mode selection scheme for an intelligent reflecting surface (IRS)assisted buffer-aided relaying protocol, capable of boosting throughput and reducing delay in a two-hop wireless network. The IRS is composed of multiple passive elements and acts as a low-cost multi-antenna relay node, which is capable of relaying (reflecting) an incident signal to the receiver in a full-duplex manner. The proposed protocol has multiple relay activation modes, such as an IRS-assisted full-duplex mode, a virtual fullduplex mode, and multiple half-duplex modes, and constitutes a unified full-duplex transmission. Our performance results demonstrate that the proposed IRS-assisted unified full-duplex buffer-aided relaying transmission outperforms the benchmark schemes.