Visual Patterns Research Articles

Fast, Simple, and Accurate Time Series Analysis with Large Language Models: An Example of Mean-motion Resonances Identification

Classical machine learning has been actively utilized in astronomy to address various challenges, including predicting orbital stability, classifying asteroids, galaxies, and other objects, and analyzing images. However, the emerging trend in artificial intelligence involves the use of large language models such as GPT-4 and ChatGPT. These models are trained on a large corpus of text and can perform a wide range of natural language processing tasks, including text generation, translation, summarization, and classification. Surprisingly, these capabilities present significant potential for application in astronomy. This paper demonstrates how the new model gpt-4-vision-preview can analyze visual patterns and accurately classify asteroids as resonant or nonresonant with high accuracy. This process requires no training, fine-tuning, or coding beyond writing the appropriate prompt in natural language. Moreover, this approach can be extended to other common problems within astronomy.

Pruning Self-Attentions Into Convolutional Layers in Single Path.

Vision Transformers (ViTs) have achieved impressive performance over various computer vision tasks. However, modeling global correlations with multi-head self-attention (MSA) layers leads to two widely recognized issues: the massive computational resource consumption and the lack of intrinsic inductive bias for modeling local visual patterns. To solve both issues, we devise a simple yet effective method named Single-Path Vision Transformer pruning (SPViT), to efficiently and automatically compress the pre-trained ViTs into compact models with proper locality added. Specifically, we first propose a novel weight-sharing scheme between MSA and convolutional operations, delivering a single-path space to encode all candidate operations. In this way, we cast the operation search problem as finding which subset of parameters to use in each MSA layer, which significantly reduces the computational cost and optimization difficulty, and the convolution kernels can be well initialized using pre-trained MSA parameters. Relying on the single-path space, we introduce learnable binary gates to encode the operation choices in MSA layers. Similarly, we further employ learnable gates to encode the fine-grained MLP expansion ratios of FFN layers. In this way, our SPViT optimizes the learnable gates to automatically explore from a vast and unified search space and flexibly adjust the MSA-FFN pruning proportions for each individual dense model. We conduct extensive experiments on two representative ViTs showing that our SPViT achieves a new SOTA for pruning on ImageNet-1 k. For example, our SPViT can trim 52.0% FLOPs for DeiT-B and get an impressive 0.6% top-1 accuracy gain simultaneously.

Deepfake Detection System

Abstract: The emergence of deepfake technology presents a profound challenge to the integrity and trustworthiness of multimedia content online. To address this issue, this study proposes a novel deepfake detection system that integrates a hybrid Recurrent Neural Network (RNN), Convolutional Neural Network (CNN), and Long Short-Term Memory (LSTM) architecture. The proposed system employs a multi-faceted approach to training by utilizing several diverse datasets encompassing real and synthetic videos. This comprehensive training strategy ensures that the model learns robust features representative of both authentic and manipulated content across various contexts and scenarios. By incorporating a range of datasets, including those specifically curated to represent different deepfake generation techniques and quality levels, the model gains a more comprehensive understanding of the intricate nuances present in manipulated videos. During the training phase, the CNN component extracts high-level spatial features from individual frames of the input videos, effectively capturing visual patterns indicative of deepfake manipulation.

A multi-resolution convolutional attention network for efficient diabetic retinopathy classification

Diabetic retinopathy (DR) is a disorder commonly associated with individuals who have diabetes and is caused by excessive glucose levels in the blood, which induce retinal damage. Manual diagnosis of DR by medical experts is time-consuming; hence, a computer-aided diagnosis (CAD) solution for faster identification and improved efficiency is required. Diagnosis of DR is complex due to the prevalence of inter-class differences and small lesions, making it difficult for convolutional neural networks (CNNs) to identify discriminative regions effectively. The most critical challenge is locating complex patterns connected with visual differences and lesion locations. To tackle the aforementioned challenges, this work presents a novel multi-resolution convolutional attention network (MuR-CAN), which improves overall performance by emphasizing discriminative features. MuR-CAN includes a multi-dilation attention block (MDAB), which uses depth-wise convolution layers with various dilation rates to allow the model to prioritize relevant features and capture multi-scale spatial information. This method provides for simultaneously highlighting fine features and subtle visual patterns. Furthermore, the features extracted from the MuR-CAN are utilized to train the support vector machine (SVM) for DR classification. The integration of the SVM classifier with the proposed CNN architecture ensures improved accuracy. The experimental findings demonstrate that the proposed model outperforms the current state-of-the-art methods.

Driving patterns in connected environments: A case study of intersection-approaching behavior of professional and non-professional drivers

The in-vehicle communication provides promising opportunities to improve the road safety and traffic efficiency. Previous studies demonstrated that the professional drivers have better driving skills than the non-professional drivers who allocate more attention to secondary tasks. However, they may not be sensitive to the new in-vehicle technology. In addition, these qualitative studies failed to elaborate on the visual and response behavior differences among different driver groups (professional drivers such as taxi, bus, motorcoach, and non-professional drivers), and lacked the quantitative analysis of driving patterns in a new environment. This paper explores the differences in visual interaction, response characteristics, driving performance, and behavior patterns between the professional and non-professional drivers in the connected environment through a case study of intersection-approaching behavior using a driving simulator. More precisely, two driving scenarios (baseline and human–machine interface (HMI)) were designed in the driving simulator, and 65 participants, including 34 professional drivers and 31 non-professional drivers, completed the experiment. In the HMI scenario, the driver was provided with the signal light phase and phase transition remaining time of the current intersection. This paper also proposes a driving pattern extraction model based on the Bayesian non-parametric method combined with a text clustering algorithm to perform a quantitative description of the driving patterns. The results show that the professional drivers tend to interact less with the HMI compared with the non-professional drivers. Moreover, the professional drivers’ first gaze at the HMI occurs and responds earlier. The proposed driving model can effectively describe 7 patterns of intersection-approaching behavior. The connected information can significantly improve the efficiency of the intersection traffic and the driving behavior. However, the professional drivers are more responsive and behave more consistently. This study can provide insights into the development of personalized assisted driving systems, as the two driving populations differ in their interactions, responses, and behavioral patterns.

WEB VISUALIZATION AND SPATIAL ANALYSIS OF TREE SPECIES DISTRIBUTION AT UNIVERSITY OF THE PHILIPPINES DILIMAN ACADEMIC OVAL

Abstract. Understanding the distribution of tree species is crucial for ecological comprehension and effective conservation strategies. This research presents a web-based platform aimed at visually representing the trees found at the University of the Philippines Diliman campus. Open-source geospatial tools such as OpenStreetMap, KoboToolbox, Pl@ntNet, JOSM, uMap, and QGIS were employed for data visualization, collection, analysis, validation, and spatial pattern assessment. Through these tools, a total of 11 tree species were identified, with acacia emerging as the dominant species, especially concentrated along the academic oval. The application of Kernel Density Estimation (HeatMap) in QGIS revealed these spatial distribution patterns. The findings contribute to the effective management, conservation, and restoration of the campus' green spaces, shedding light on the significance of trees in urban ecosystems. The online interactive map developed serves as a valuable tool for improving management and conservation strategies. Importantly, the study assesses the effectiveness of open-source tools, showcasing their versatility and comparable capabilities to proprietary alternatives in geospatial applications.

Migraine aura discrimination using machine learning: an fMRI study during ictal and interictal periods.

Functional magnetic resonance imaging (fMRI) studies on migraine with aura are challenging due to the rarity of patients with triggered cases. This study optimized methodologies to explore differences in ictal and interictal spatiotemporal activation patterns based on visual stimuli using fMRI in two patients with unique aura triggers. Both patients underwent separate fMRI sessions during the ictal and interictal periods. The Gaussian Process Classifier (GPC) was used to differentiate these periods by employing a machine learning temporal embedding approach and spatiotemporal activation patterns based on visual stimuli. When restricted to visual and occipital regions, GPC had an improved performance, with accuracy rates for patients A and B of roughly 86-90% and 77-81%, respectively (p < 0.01). The algorithm effectively differentiated visual stimulation and rest periods and identified times when aura symptoms manifested, as evident from the varying predicted probabilities in the GPC models. These findings contribute to our understanding of the role of visual processing and brain activity patterns in migraine with aura and the significance of temporal embedding techniques in examining aura phenomena. This finding has implications for diagnostic tools and therapeutic techniques, especially for patients suffering from aura symptoms.

Towards a typology of virtual tourists: Efficacy of visual patterns and attentional cues

Virtual tourism is rapidly emerging in prominence in contemporary tourism research. Previous studies tend to consider the virtual tourism population as a homogeneous entity, resulting in unintended consequences. Despite the ocular-centric nature within virtual environments, existing eye tracking studies have rarely classified virtual tourists into benefit-based segments, discounting the critical role of pre-existing visual patterns and associated attentional cues. Subsequently, this research constructs a theoretically rigorous and evidence-informed typology of virtual tourists, based on attention patterns discerned from eye tracking trajectory and associated visual cues, triangulated with interview data. Building on the theory of bottom-up and top-down attentional processing, four discrete virtual tourist typologies emerged in terms of visual completeness and visual cognition: visual passives, curious scanners, focused observers, and comprehensive engagers. The visual and cognitive characteristics among the groups were collated, revealing significant differences articulated between each of the virtual tourist typologies. This research not only advances knowledge through a renewed emphasis on typology-driven theorising within the virtual tourism population, but also offers practical insights into implementing an innovative and effective segmentation approach. Future research should consider comparing the theoretical and empirical specifications of the virtual tourist typology.

Decoding Immuno-Competence: A Novel Analysis of Complete Blood Cell Count Data in COVID-19 Outcomes.

While 'immuno-competence' is a well-known term, it lacks an operational definition. To address this omission, this study explored whether the temporal and structured data of the complete blood cell count (CBC) can rapidly estimate immuno-competence. To this end, one or more ratios that included data on all monocytes, lymphocytes and neutrophils were investigated. Longitudinal CBC data collected from 101 COVID-19 patients (291 observations) were analyzed. Dynamics were estimated with several approaches, which included non-structured (the classic CBC format) and structured data. Structured data were assessed as complex ratios that capture multicellular interactions among leukocytes. In comparing survivors with non-survivors, the hypothesis that immuno-competence may exhibit feedback-like (oscillatory or cyclic) responses was tested. While non-structured data did not distinguish survivors from non-survivors, structured data revealed immunological and statistical differences between outcomes: while survivors exhibited oscillatory data patterns, non-survivors did not. In survivors, many variables (including IL-6, hemoglobin and several complex indicators) showed values above or below the levels observed on day 1 of the hospitalization period, displaying L-shaped data distributions (positive kurtosis). In contrast, non-survivors did not exhibit kurtosis. Three immunologically defined data subsets included only survivors. Because information was based on visual patterns generated in real time, this method can, potentially, provide information rapidly. The hypothesis that immuno-competence expresses feedback-like loops when immunological data are structured was not rejected. This function seemed to be impaired in immuno-suppressed individuals. While this method rapidly informs, it is only a guide that, to be confirmed, requires additional tests. Despite this limitation, the fact that three protective (survival-associated) immunological data subsets were observed since day 1 supports many clinical decisions, including the early and personalized prognosis and identification of targets that immunomodulatory therapies could pursue. Because it extracts more information from the same data, structured data may replace the century-old format of the CBC.

Model compression of deep neural network architectures for visual pattern recognition: Current status and future directions

Visual Pattern Recognition Networks (VPRNs) are widely used in various visual data based applications such as computer vision and edge AI. VPRNs help to enhance a machine's learning by extracting underlying patterns from input images and videos. VPRNs can be implemented using statistical, syntactical, or Machine Learning (ML) based approaches. With the popularity of Deep Neural Networks (DNNs) in the recent past, DNN based VPRNs have become an inevitable choice due to their suitability for handling such high dimensional data.However, such DNN based VPRNs bring along a curse of dimensionality, leading to intricate computations, substantial memory demands, and increased energy requirements. This impedes their practical deployment in resource constrained and strict latency required environments. Such overheads impose a demand for compression of VPRNs without impairing their performance. This research presents an exhaustive survey on compression techniques used for sequential, non-sequential and advanced DNN architectures from the perspective of Visual Pattern Recognition (VPR). Research findings of this study reveal potential scope to compress non-sequential, sequential and advanced DNN architectures for VPR applications. The study also clarifies challenges, reported and unreported issues as well as research gaps and provides directions for further research in this domain.

Capnography waveforms: basic interpretation in neonatal intensive care.

End-tidal capnography can provide useful clinical information displayed on the ventilator screen or bedside monitor. It is important that clinicians can assess and utilise this information to assist in identifying underlying complications and pulmonary pathology. Sudden change or loss of the CO2 waveform can act as a safety measure in alerting clinicians of a dislodged or blocked endotracheal tube, considering the concurrent flow and volume waveforms. Visual pattern recognition by the clinicians of commonly seen waveform traces may act as an adjunct to other modes of ventilatory monitoring techniques. Waveforms traces can aid clinical management, help identify cases of ventilation asynchrony between the infant and the ventilator. We present some common clinical scenarios where tidal capnography can be useful in the timely identification of pulmonary complication and for practical troubleshooting at the cot-side.

The function of web decorations in orb web spiders

Visual signal, mechanical reinforcement, protein storage, or non-functional stress response? Web decorations constructed by a number of orb web spider species puzzle behavioural ecologists. While some species use a variety of construction materials, it seems particularly difficult to solve the issue for silk decorations. The visual pattern of decoration structures has sparked the researchers’ imagination, and the conclusion that they act as signals is unsurprising. For over a century, however, we have not found a conclusive answer on a specific signal function of silk decorations. More recent studies even suggest that the construction mode of web decorations may render a specific signal function rather unlikely. In this review, I discuss reasons for the continuous struggle to find conclusive answers and what could be alternative routes for unravelling their adaptive significance. Based on my own experience in the field, I present a personal viewpoint, which I hope will be inspiring at a stage where research in this field seems to have reached a dead end. We are faced with a controversial debate, inconclusive and sometimes contradicting results; and an interest in new studies is fading. I draw the attention to three trouble areas, covering research gaps, logical inconsistencies and conceptual misunderstandings. More specifically: 1) Web decorations should be treated more as the dependent variable. 2) Experimental setups of several earlier studies appear flawed in retrospect, and their results thus overinterpreted. 3) We have not understood the evolutionary origin of web decoration. We may look at a signal that is still in an early phase of shaping, and inconclusive results may thus be inevitable. Finally, if web decorations do not act as signals, studies that look for exactly that cannot find conclusive results. In discussing these issues critically, I argue, we can open new routes for progress in finding a satisfying solution of the riddle of the silk decoration function.

The theatrics of Arabic script: word and/as image in the diagrams of the Kitāb al-diryāq (BnF arabe 2964)

This article focuses upon the arrangement of written words in geometric shapes or patterns—specifically in Arabic script. It takes the case of the Kitāb al-diryāq (Book of Antidotes, 595 AH/1199 CE, BnF arabe 2964), an illuminated and illustrated manuscript that features a series of geometric compositions created using Arabic script. The script is employed in conjunction with or independent of geometric lines and curves, with its ligatures compressed and elongated, and words repeated, ornamented, or oriented multidirectionally to augment the geometric shape. The aesthetic manipulation of the script in this manner severely compromises the legibility of the words—the pages invite a viewing of the script rather than a reading of the text. By closely examining a set of four such compositions—articulated as diagrams—the article demonstrates the semantic value of the visual script patterns when read in concert with the verbal content. Furthermore, by focusing on the extralinguistic components of the diagrams, I argue that the arrangement of text in geometric patterns is a vital medium of transmitting knowledge—the form not only shapes the reception of verbal messages but also augments an aesthetic image; operates as a cognitive device; and transforms the text into a scientific representation. The diagrams visually and semantically shape knowledge.

Padang Cuisine Classification using Deep Convolutional Neural Networks (CNN) and Transfer Learning

Padang cuisine, as an integral part of Indonesia's culinary wealth, challenges researchers to explore its potential through technological approaches. This article explores the application of Deep Convolutional Neural Networks (DCNN) and Transfer Learning for the classification of Padang cuisine. DCNN is used to understand and recognize unique visual patterns in food images, while Transfer Learning leverages existing knowledge to improve the model's performance in the specific task of classifying Padang cuisine. This process involves collecting a representative dataset, data preprocessing, and model training to create a system capable of accurately identifying various Padang dishes. Experimental results show that this approach achieves an accuracy rate of 92%, demonstrating its effectiveness in classifying different types of Padang cuisine. By engaging advanced technology, this article contributes to a deeper understanding of the integration between traditional culinary richness and innovation in the modern world of artificial intelligence. These findings indicate that AI implementation can be used to document and celebrate diverse culinary heritage, as well as support the culinary industry in managing and identifying dishes in real-time.

Classification of types Roasted Coffee Beans using Convolutional Neural Network Method

In the current digital era, the role of technology in the agricultural industry is very necessary to increase yields which can have an impact on the productivity and welfare of farmers. Coffee is a drink that has been very popular for many years. Due to the high demand for coffee beans, this research aims to develop a system that can classify types of roasted coffee beans based on images using the Convolution Neural Network (CNN) method. Coffee bean processing is the most important stage in the coffee industry, classifying coffee beans often requires more in-depth knowledge and extensive experience regarding coffee beans. Therefore, this system can be a more effective solution. The author collects a dataset containing types of roasted coffee beans, then the Convolutional Neural Network (CNN) can analyze in the form of visual patterns each type of coffee bean. This implementation is expected to help the coffee industry identify coffee beans quickly and accurately.

Brain Tumor Detection Using Classification Model

This study aims to develop automated methods for detecting and segmenting brain tumors in MRI scans using neural networks and conventional image processing. Accurately identifying tumor location and boundaries is crucial for diagnosis and determining the most effective treatment approach. Manual review by radiologists is time-consuming and subjective. However, recent advances in deep learning now enable more efficient analysis of large MRI datasets.&#x0D; A variety of neural network models have been explored for tumor segmentation, including recurrent networks, autoencoders and convolutional networks. RNNs are well-suited to capture contextual information across sequential MRI slices. Autoencoders perform unsupervised feature learning from the images. CNNs have shown promise due to their ability to directly learn visual patterns from pixel values. Unsupervised clustering techniques such as fuzzy c-means and k-means are also investigated to group tumor pixels based on intensity similarity without labeled data.&#x0D; Previous research combining these techniques has achieved high segmentation accuracy between 84-97% using additional image processing steps. Automated segmentation provides benefits over manual inspection such as reduced workload and more consistent delineation of tumor margins. It can precisely outline growths to aid radiologists in diagnosis and pathologists in assessing malignancy.&#x0D; The detection and segmentation of brain tumors has applications in both radiology and oncology. Radiologists could benefit from faster scan review and automated report generation. Neurosurgeons would gain insights into a tumor's size, location and infiltration to determine the most suitable surgical approach. Oncologists could utilize quantitative tumor features to select the most targeted radiation therapy or chemotherapy protocols. Overall, continued advances in deep and machine learning hold promise to improve diagnosis and management of brain cancer patients.

From Mind-Reading to Health-Reading Machines: Towards Contactless Health Diagnosis using Generative Artificial Intelligence

Generative Artificial Intelligence (AI) has recently shown impressive results across a wide spectrum of tasks. Inspired by the emerging notion of mind-reading machines and by the medical evidences indicating that some diseases and brain disorders produce facial abnormalities and interrupt normal facial expression formation, we are (1) establishing, in close interaction with medical practitioners, the basic understanding of the correlations between visual patterns and health conditions, and (2) developing novel computational models, based on Generative AI, to learn the identified correlations. Devising computational models for detecting abnormalities reflective of internal diseases based only on visual information is a highly intriguing research problem. The goal is to achieve ground-breaking results in unobtrusive health diagnosis and to gain insight into the correlation between face, body, and mind. This work has the potential to open new research directions in several fields and to bridge the gap between the research activities conducted in several disciplines like engineering and medicine. This article explains the main idea behind contactless health reading using generative artificial intelligence.

The Effect of Different Visual Literacy Patterns of Drawings and Instructional Photos in Developing Concepts for Middle School in State of Kuwait

The Effect of Different Visual Literacy Patterns of Drawings and Instructional Photos in Developing Concepts for Middle School in State of Kuwait

The incidence and prevalence of diabetic macular edema and proliferative diabetic retinopathy, their progression to visual impairment and patterns in their intravitreal treatment in the Finnish population.

The worldwide prevalence of diabetes mellitus (DM) continues to increase. As DM is linked to various ophthalmological comorbidities, it is crucial to understand the incidence and the treatment patterns of these complications to minimise the treatment burden for the patient and the healthcare system. This study aims to evaluate the incidence and prevalence of diabetic macular oedema (DME) and proliferative diabetic retinopathy (PDR) and to analyse intravitreal (IVT) treatment patterns and responses in the Finnish population with diabetes. A nationwide data register containing details of over 20-year-old individuals with diabetes was used in the analyses. The incidence and prevalence of DME and PDR among the Finnish population with diabetes either declined or remained stable during 2007-2017 (Incidence rate: DME -40.8%, PDR -65.3%; prevalence rate: DME +4.7%, PDR -11.2%). During the same period, number of persons suffering from diabetes increased by +58.3%. The total number of IVT injections increased by 261.7%; the number of patients receiving IVT treatments increased by 133.6% from 2011 to 2017, reflecting changes in patient numbers in the ophthalmology departments. Furthermore, irrespective of the rising number of patients with diabetes, the numbers with visual impairment declined by 75.8% among DME and by 75.7% among PDR patients in 2007-2017. Regardless of the considerable increase in the workload of ophthalmology departments, the healthcare system has been able to reduce both the age and sex standardised incidence of DME and PDR among the diabetic population suffering from a visual impairment associated with this disease.

Decoupled Contrastive Learning for Long-Tailed Recognition

Supervised Contrastive Loss (SCL) is popular in visual representation learning. Given an anchor image, SCL pulls two types of positive samples, i.e., its augmentation and other images from the same class together, while pushes negative images apart to optimize the learned embedding. In the scenario of long-tailed recognition, where the number of samples in each class is imbalanced, treating two types of positive samples equally leads to the biased optimization for intra-category distance. In addition, similarity relationship among negative samples, that are ignored by SCL, also presents meaningful semantic cues. To improve the performance on long-tailed recognition, this paper addresses those two issues of SCL by decoupling the training objective. Specifically, it decouples two types of positives in SCL and optimizes their relations toward different objectives to alleviate the influence of the imbalanced dataset. We further propose a patch-based self distillation to transfer knowledge from head to tail classes to relieve the under-representation of tail classes. It uses patch-based features to mine shared visual patterns among different instances and leverages a self distillation procedure to transfer such knowledge. Experiments on different long-tailed classification benchmarks demonstrate the superiority of our method. For instance, it achieves the 57.7% top-1 accuracy on the ImageNet-LT dataset. Combined with the ensemble-based method, the performance can be further boosted to 59.7%, which substantially outperforms many recent works. Our code will be released.