scholarly journals Artwork Style Recognition Using Vision Transformers and MLP Mixer

Technologies ◽  
2021 ◽  
Vol 10 (1) ◽  
pp. 2
Author(s):  
Lazaros Alexios Iliadis ◽  
Spyridon Nikolaidis ◽  
Panagiotis Sarigiannidis ◽  
Shaohua Wan ◽  
Sotirios K. Goudos
Keyword(s):  
Prediction Accuracy ◽  
Extensive Study ◽  
Multi Layer Perceptron ◽  
Future Studies ◽  
Artistic Style ◽  
Recurrent Processing ◽  
Rich Information ◽  
Learning Architectures ◽  
Dominant Paradigm

Through the extensive study of transformers, attention mechanisms have emerged as potentially more powerful than sequential recurrent processing and convolution. In this realm, Vision Transformers have gained much research interest, since their architecture changes the dominant paradigm in Computer Vision. An interesting and difficult task in this field is the classification of artwork styles, since the artistic style of a painting is a descriptor that captures rich information about the painting. In this paper, two different Deep Learning architectures—Vision Transformer and MLP Mixer (Multi-layer Perceptron Mixer)—are trained from scratch in the task of artwork style recognition, achieving over 39% prediction accuracy for 21 style classes on the WikiArt paintings dataset. In addition, a comparative study between the most common optimizers was conducted obtaining useful information for future studies.

Values in Action Inventory of Strengths (VIA-IS): Translation and Validation in Urdu Language

2020 ◽  
Vol 35 (1) ◽  
pp. 163-189
Author(s):  
Afifa Anjum ◽  
Naumana Amjad
Keyword(s):  
Factor Structure ◽  
Convergent Validity ◽  
Character Strengths ◽  
Future Studies ◽  
Structure Factors ◽  
Adult Participants ◽  
Positive Correlations ◽  
Cross Language ◽  
Translation And Validation

Values in Action is a classification of 24 character strengths grouped under six virtue categories. This classification is claimed to be universal across cultures and religions (Peterson & Seligman, 2004) and its measure that is, Values in Action Inventory of Strengths (VIA-IS) has been translated and validated in many languages. The present study aimed at its Urdu translation and validation on Pakistani adults taken from different educational institutes and workplaces. Study comprised two parts. Part I dealt with the translation and cross-language validation while in Part II, Construct validation on a sample of 542 adults and convergent validity on a sample of 210 adult participants were determined. Findings revealed satisfactory alpha coefficients for Urdu version. Significant positive correlations with positive affect and life satisfaction and negative correlations with negative affect were indicators of its convergent validity. Age was negatively associated with five strengths whereas significant gender differences were found on seven strengths. Social desirability effects were nonsignificant. Strength-to-virtue level factor structure exploration resulted in a theoretically meaningful four factor structure. Factors were named as Interpersonal, Cognitive, Vitality, and Transcendence and were comparable to factor structures proposed in studies on VIA-IS from a few other cultures. The study offers a valid Urdu translation for use in future studies with adult Urdu speaking population.

scholarly journals A Machine Learning Approach to Study Glycosidase Activities from Bifidobacterium

2021 ◽  
Vol 9 (5) ◽  
pp. 1034
Author(s):  
Carlos Sabater ◽  
Lorena Ruiz ◽  
Abelardo Margolles
Keyword(s):  
Machine Learning ◽  
Supervised Classification ◽  
Machine Learning Algorithms ◽  
Learning Approach ◽  
Human Milk Oligosaccharides ◽  
Future Studies ◽  
High Fiber ◽  
Machine Learning Approach ◽  
Prebiotic Oligosaccharides

This study aimed to recover metagenome-assembled genomes (MAGs) from human fecal samples to characterize the glycosidase profiles of Bifidobacterium species exposed to different prebiotic oligosaccharides (galacto-oligosaccharides, fructo-oligosaccharides and human milk oligosaccharides, HMOs) as well as high-fiber diets. A total of 1806 MAGs were recovered from 487 infant and adult metagenomes. Unsupervised and supervised classification of glycosidases codified in MAGs using machine-learning algorithms allowed establishing characteristic hydrolytic profiles for B. adolescentis, B. bifidum, B. breve, B. longum and B. pseudocatenulatum, yielding classification rates above 90%. Glycosidase families GH5 44, GH32, and GH110 were characteristic of B. bifidum. The presence or absence of GH1, GH2, GH5 and GH20 was characteristic of B. adolescentis, B. breve and B. pseudocatenulatum, while families GH1 and GH30 were relevant in MAGs from B. longum. These characteristic profiles allowed discriminating bifidobacteria regardless of prebiotic exposure. Correlation analysis of glycosidase activities suggests strong associations between glycosidase families comprising HMOs-degrading enzymes, which are often found in MAGs from the same species. Mathematical models here proposed may contribute to a better understanding of the carbohydrate metabolism of some common bifidobacteria species and could be extrapolated to other microorganisms of interest in future studies.

scholarly journals The Classification of Medicinal Plant Leaves Based on Multispectral and Texture Feature Using Machine Learning Approach

Agronomy ◽  
2021 ◽  
Vol 11 (2) ◽  
pp. 263
Author(s):  
Samreen Naeem ◽  
Aqib Ali ◽  
Christophe Chesneau ◽  
Muhammad H. Tahir ◽  
Farrukh Jamal ◽  
...  
Keyword(s):  
Machine Learning ◽  
Medicinal Plant ◽  
Texture Feature ◽  
Stevia Rebaudiana ◽  
Ocimum Sanctum ◽  
Multi Layer Perceptron ◽  
Plant Leaves ◽  
Chi Square ◽  
Lemon Balm

This study proposes the machine learning based classification of medical plant leaves. The total six varieties of medicinal plant leaves-based dataset are collected from the Department of Agriculture, The Islamia University of Bahawalpur, Pakistan. These plants are commonly named in English as (herbal) Tulsi, Peppermint, Bael, Lemon balm, Catnip, and Stevia and scientifically named in Latin as Ocimum sanctum, Mentha balsamea, Aegle marmelos, Melissa officinalis, Nepeta cataria, and Stevia rebaudiana, respectively. The multispectral and digital image dataset are collected via a computer vision laboratory setup. For the preprocessing step, we crop the region of the leaf and transform it into a gray level format. Secondly, we perform a seed intensity-based edge/line detection utilizing Sobel filter and draw five regions of observations. A total of 65 fused features dataset is extracted, being a combination of texture, run-length matrix, and multi-spectral features. For the feature optimization process, we employ a chi-square feature selection approach and select 14 optimized features. Finally, five machine learning classifiers named as a multi-layer perceptron, logit-boost, bagging, random forest, and simple logistic are deployed on an optimized medicinal plant leaves dataset, and it is observed that the multi-layer perceptron classifier shows a relatively promising accuracy of 99.01% as compared to the competition. The distinct classification accuracy by the multi-layer perceptron classifier on six medicinal plant leaves are 99.10% for Tulsi, 99.80% for Peppermint, 98.40% for Bael, 99.90% for Lemon balm, 98.40% for Catnip, and 99.20% for Stevia.

Classification of vegetation communities in which geese rear broods on the Yukon – Kuskokwim delta, Alaska

1994 ◽  
Vol 72 (9) ◽  
pp. 1294-1301 ◽  
Author(s):  
Christopher A. Babcock ◽  
Craig R. Ely
Keyword(s):  
Salt Marsh ◽  
Plant Communities ◽  
Indicator Species Analysis ◽  
Future Studies ◽  
Brood Rearing ◽  
Species Analysis ◽  
Twinspan Classification ◽  
Vegetation Complexes ◽  
Air Photo

Plant communities are described from an area on the Yukon – Kuskokwim (Y-K) delta of Alaska that is used extensively for brood rearing by three species of geese. Earlier studies identified plant species important as food for young geese, but few studies describe or quantify plant communities. We classified species presence or absence information from over 700 quadrats using a two-way indicator species analysis (TWINSPAN) and then tested for agreement of signatures on colour infrared air photos with the identified communities. Sedges were found to dominate all but the wettest and driest communities. Most of the brood-rearing area was covered by Carex ramenskii and Carex rariflora meadows, ponds, Carex mackenziei-dominated pond margins, and C. ramenskii and grass levee meadows. Our interpretation of airphotos accurately predicted vegetation community classes, which will facilitate future studies of habitat selection by geese during the time they are rearing young. The TWINSPAN classification was comparable to classifications of studies conducted elsewhere on the Y-K delta. The interpretation of air photos will enable the identification and evaluation of wetland vegetation complexes and potential goose brood-rearing areas away from our study site. Key words: air-photo interpretation, Alaska, plant communities, salt marsh, Yukon – Kuskokwim delta.

A GALLERY OF CHUA ATTRACTORS PART VI

2007 ◽  
Vol 17 (06) ◽  
pp. 1801-1910 ◽  
Author(s):  
ELEONORA BILOTTA ◽  
GIANPIERO DI BLASI ◽  
FAUSTO STRANGES ◽  
PIETRO PANTANO
Keyword(s):  
Phase Space ◽  
Spatial Organization ◽  
Fractal Dimensions ◽  
Point Of View ◽  
Experimental Investigations ◽  
Technical Data ◽  
Future Studies ◽  
Small Set ◽  
Methodological Approaches

In this article, we conclude our series of papers on the analysis and visualization of Chua attractors and their generalizations. We present a gallery of 144 n-scroll, 15 hyperchaotic and 37 synchronized systems. Along with time series and FFT we provide 3D visualizations; for some attractors we also supply Lyapunov coefficients and fractal dimensions. The goal in constructing our Gallery has been to make the general public aware of the enormous variety of chaotic phenomena and to change the widespread impression that they are isolated rarities. The Gallery provides a valuable collection of images and technical data which can be used to analyze these phenomena and to reproduce them in future studies. From a scientific point of view, we have tried to identify new methodological approaches to the study of chaos, opening nontraditional perspectives on the complexity of this domain. In our papers, we have discussed a broad range of topics, ranging from techniques for visualizing Chua attractors to computational methods allowing us to make a statistical classification of attractors' positions in phase space and to describe the evolutionary processes through which their shapes change over time. We see these processes as analogous to population dynamics in artificial environments. Within these environments, we use experimental methods to identify the models which guide morphogenetic change and which organize genetic landscapes in parameter space. This paper is organized as follows. First, we provide formal descriptions of the attractors generated by n-scroll, hyperchaotic and synchronized systems. The next section describes a Gallery of Chua attractors, generated by gradually varying the parameters and analyzing the resulting bifurcation maps. We then describe software tools allowing us to perform statistical analyses on selected sets of attractors, to visualize them, to explore their organization in phase space, and to conduct experimental investigations of the morphogenetic processes through which a small set of base attractors can generate a broad range of different forms. In the last section, we describe the creation of a Virtual 3D Gallery displaying some of the attractors we have presented in our six papers. The attractors are organized by theme, as they might be in a museum. The environment allows users to explore the attractors, interact with shapes, listen to music and sounds generated by the attractors, change their spatial organization, and create new shapes. To complete the paper — and the series — we propose a number of general conclusions.

A Study on Fault Classification of Solenoid Pumps based on Multi-Layer Perceptron

2021 ◽  
Vol 21 (1) ◽  
pp. 12-19
Author(s):  
Suju Kim ◽  
Akpudo Ugochukwu Ejike ◽  
Jangwook Hur
Keyword(s):  
Fault Classification ◽  
Multi Layer Perceptron

How much do periventricular lesions assist in distinguishing migraine with aura from CIS?

Neurology ◽  
2019 ◽  
Vol 92 (15) ◽  
pp. e1739-e1744 ◽  
Author(s):  
Caterina Lapucci ◽  
Laura Saitta ◽  
Giulia Bommarito ◽  
Maria Pia Sormani ◽  
Matteo Pardini ◽  
...  
Keyword(s):  
Multiple Sclerosis ◽  
Migraine With Aura ◽  
Diagnostic Performance ◽  
White Matter Hyperintensities ◽  
Clinically Isolated Syndrome ◽  
Lower Sensitivity ◽  
Resonance Imaging ◽  
Future Studies ◽  
Periventricular Lesions

ObjectiveTo evaluate in clinically isolated syndrome (CIS) and migraine with aura (MA) how the number of periventricular lesions (PVLs) detected at MRI influences diagnostic performance when the Magnetic Resonance Imaging in Multiple Sclerosis (MAGNIMS) or the 2017 revised criteria are applied.MethodsIn this retrospective study, white matter hyperintensities (WMH) of 84 patients with MA and 79 patients with CIS were assessed using manual segmentation technique. Lesion probability maps (LPMs) and voxel-wise analysis of lesion distribution by diagnosis were obtained. Furthermore, we performed a logistic regression analysis based on lesion locations and volumes.ResultsCompared to patients with MA, patients with CIS showed a significant overall higher T2 WMH mean number and volume (17.9 ± 16.9 vs 6.2 ± 11.9 and 3.1 ± 4.2 vs 0.3 ± 0.6 mL; p < 0.0001) and a significantly higher T2 WMH mean number in infratentorial, periventricular, and juxtacortical areas (p < 0.0001). LPMs identified the periventricular regions as the sites with the highest probability of detecting T2 WMH in patients with CIS. Voxel-wise analysis of lesion distribution by diagnosis revealed a statistically significant association exclusively between the diagnosis of CIS and the PVLs. MAGNIMS criteria demonstrated the highest specificity in differentiating patients with CIS from patients with MA (100% vs 87%) against a predictable lower sensitivity (63% vs 72%).ConclusionsPVLs play a key role in the differential diagnosis between MA and CIS, particularly when there are more than 3. Future studies on multiple sclerosis criteria might reconsider the 3 PVLs to minimize the risk of misdiagnosis.Classification of evidenceThis study provides Class IV evidence that the presence at least 3 PVLs increases the specificity in distinguishing MA from CIS.

Sign in / Sign up

Export Citation Format

Share Document