To see the other types of publications on this topic, follow the link: Additive Angular Margin loss.
Journal articles on the topic 'Additive Angular Margin loss'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 27 journal articles for your research on the topic 'Additive Angular Margin loss.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse journal articles on a wide variety of disciplines and organise your bibliography correctly.
1
Zhou, Shengwei, Caikou Chen, Guojiang Han, and Xielian Hou. "Double Additive Margin Softmax Loss for Face Recognition." Applied Sciences 10, no. 1 (December 19, 2019): 60. http://dx.doi.org/10.3390/app10010060.
Full textAbstract:
Learning large-margin face features whose intra-class variance is small and inter-class diversity is one of important challenges in feature learning applying Deep Convolutional Neural Networks (DCNNs) for face recognition. Recently, an appealing line of research is to incorporate an angular margin in the original softmax loss functions for obtaining discriminative deep features during the training of DCNNs. In this paper we propose a novel loss function, termed as double additive margin Softmax loss (DAM-Softmax). The presented loss has a clearer geometrical explanation and can obtain highly discriminative features for face recognition. Extensive experimental evaluation of several recent state-of-the-art softmax loss functions are conducted on the relevant face recognition benchmarks, CASIA-Webface, LFW, CALFW, CPLFW, and CFP-FP. We show that the proposed loss function consistently outperforms the state-of-the-art.
2
Wang, Xiaobo, Shifeng Zhang, Shuo Wang, Tianyu Fu, Hailin Shi, and Tao Mei. "Mis-Classified Vector Guided Softmax Loss for Face Recognition." Proceedings of the AAAI Conference on Artificial Intelligence 34, no. 07 (April 3, 2020): 12241–48. http://dx.doi.org/10.1609/aaai.v34i07.6906.
Full textAbstract:
Face recognition has witnessed significant progress due to the advances of deep convolutional neural networks (CNNs), the central task of which is how to improve the feature discrimination. To this end, several margin-based (e.g., angular, additive and additive angular margins) softmax loss functions have been proposed to increase the feature margin between different classes. However, despite great achievements have been made, they mainly suffer from three issues: 1) Obviously, they ignore the importance of informative features mining for discriminative learning; 2) They encourage the feature margin only from the ground truth class, without realizing the discriminability from other non-ground truth classes; 3) The feature margin between different classes is set to be same and fixed, which may not adapt the situations very well. To cope with these issues, this paper develops a novel loss function, which adaptively emphasizes the mis-classified feature vectors to guide the discriminative feature learning. Thus we can address all the above issues and achieve more discriminative face features. To the best of our knowledge, this is the first attempt to inherit the advantages of feature margin and feature mining into a unified loss function. Experimental results on several benchmarks have demonstrated the effectiveness of our method over state-of-the-art alternatives. Our code is available at http://www.cbsr.ia.ac.cn/users/xiaobowang/.
3
Kajla, Nadeem Iqbal, Malik Muhammad Saad Missen, Muhammad Muzzamil Luqman, Mickael Coustaty, Arif Mehmood, and Gyu Sang Choi. "Additive Angular Margin Loss in Deep Graph Neural Network Classifier for Learning Graph Edit Distance." IEEE Access 8 (2020): 201752–61. http://dx.doi.org/10.1109/access.2020.3035886.
Full text
4
Liu, Wenting, Li Zhou, and Jie Chen. "Face Recognition Based on Lightweight Convolutional Neural Networks." Information 12, no. 5 (April 28, 2021): 191. http://dx.doi.org/10.3390/info12050191.
Full textAbstract:
Face recognition algorithms based on deep learning methods have become increasingly popular. Most of these are based on highly precise but complex convolutional neural networks (CNNs), which require significant computing resources and storage, and are difficult to deploy on mobile devices or embedded terminals. In this paper, we propose several methods to improve the algorithms for face recognition based on a lightweight CNN, which is further optimized in terms of the network architecture and training pattern on the basis of MobileFaceNet. Regarding the network architecture, we introduce the Squeeze-and-Excitation (SE) block and propose three improved structures via a channel attention mechanism—the depthwise SE module, the depthwise separable SE module, and the linear SE module—which are able to learn the correlation of information between channels and assign them different weights. In addition, a novel training method for the face recognition task combined with an additive angular margin loss function is proposed that performs the compression and knowledge transfer of the deep network for face recognition. Finally, we obtained high-precision and lightweight face recognition models with fewer parameters and calculations that are more suitable for applications. Through extensive experiments and analysis, we demonstrate the effectiveness of the proposed methods.
5
Sun, Jingna, Wenming Yang, Riqiang Gao, Jing-Hao Xue, and Qingmin Liao. "Inter-class angular margin loss for face recognition." Signal Processing: Image Communication 80 (February 2020): 115636. http://dx.doi.org/10.1016/j.image.2019.115636.
Full text
6
Kim, Taehyeon, Eungi Hong, and Yoonsik Choe. "Deep Morphological Anomaly Detection Based on Angular Margin Loss." Applied Sciences 11, no. 14 (July 16, 2021): 6545. http://dx.doi.org/10.3390/app11146545.
Full textAbstract:
Deep anomaly detection aims to identify “abnormal” data by utilizing a deep neural network trained on a normal training dataset. In general, industrial visual anomaly detection systems distinguish between normal and “abnormal” data through small morphological differences such as cracks and stains. Nevertheless, most existing algorithms emphasize capturing the semantic features of normal data rather than the morphological features. Therefore, they yield poor performance on real-world visual inspection, although they show their superiority in simulations with representative image classification datasets. To address this limitation, we propose a novel deep anomaly detection algorithm based on the salient morphological features of normal data. The main idea behind the proposed algorithm is to train a multiclass model to classify hundreds of morphological transformation cases applied to all the given data. To this end, the proposed algorithm utilizes a self-supervised learning strategy, making unsupervised learning straightforward. Additionally, to enhance the performance of the proposed algorithm, we replaced the cross-entropy-based loss function with the angular margin loss function. It is experimentally demonstrated that the proposed algorithm outperforms several recent anomaly detection methodologies in various datasets.
7
Li, Zhaoqun, Cheng Xu, and Biao Leng. "Angular Triplet-Center Loss for Multi-View 3D Shape Retrieval." Proceedings of the AAAI Conference on Artificial Intelligence 33 (July 17, 2019): 8682–89. http://dx.doi.org/10.1609/aaai.v33i01.33018682.
Full textAbstract:
How to obtain the desirable representation of a 3D shape, which is discriminative across categories and polymerized within classes, is a significant challenge in 3D shape retrieval. Most existing 3D shape retrieval methods focus on capturing strong discriminative shape representation with softmax loss for the classification task, while the shape feature learning with metric loss is neglected for 3D shape retrieval. In this paper, we address this problem based on the intuition that the cosine distance of shape embeddings should be close enough within the same class and far away across categories. Since most of 3D shape retrieval tasks use cosine distance of shape features for measuring shape similarity, we propose a novel metric loss named angular triplet-center loss, which directly optimizes the cosine distances between the features. It inherits the triplet-center loss property to achieve larger inter-class distance and smaller intra-class distance simultaneously. Unlike previous metric loss utilized in 3D shape retrieval methods, where Euclidean distance is adopted and the margin design is difficult, the proposed method is more convenient to train feature embeddings and more suitable for 3D shape retrieval. Moreover, the angle margin is adopted to replace the cosine margin in order to provide more explicit discriminative constraints on an embedding space. Extensive experimental results on two popular 3D object retrieval benchmarks, ModelNet40 and ShapeNetCore 55, demonstrate the effectiveness of our proposed loss, and our method has achieved state-ofthe-art results on various 3D shape datasets.
8
Chowdhury, Labib, Hasib Zunair, and Nabeel Mohammed. "Robust Deep Speaker Recognition: Learning Latent Representation with Joint Angular Margin Loss." Applied Sciences 10, no. 21 (October 26, 2020): 7522. http://dx.doi.org/10.3390/app10217522.
Full textAbstract:
Speaker identification is gaining popularity, with notable applications in security, automation, and authentication. For speaker identification, deep-convolutional-network-based approaches, such as SincNet, are used as an alternative to i-vectors. Convolution performed by parameterized sinc functions in SincNet demonstrated superior results in this area. This system optimizes softmax loss, which is integrated in the classification layer that is responsible for making predictions. Since the nature of this loss is only to increase interclass distance, it is not always an optimal design choice for biometric-authentication tasks such as face and speaker recognition. To overcome the aforementioned issues, this study proposes a family of models that improve upon the state-of-the-art SincNet model. Proposed models AF-SincNet, Ensemble-SincNet, and ALL-SincNet serve as a potential successor to the successful SincNet model. The proposed models are compared on a number of speaker-recognition datasets, such as TIMIT and LibriSpeech, with their own unique challenges. Performance improvements are demonstrated compared to competitive baselines. In interdataset evaluation, the best reported model not only consistently outperformed the baselines and current prior models, but also generalized well on unseen and diverse tasks such as Bengali speaker recognition.
9
Девисилов, Vladimir Devisilov, Шарай, and E. Sharay. "Current Stability Limits in Hydrodynamic Filter." Safety in Technosphere 2, no. 4 (August 25, 2013): 23–29. http://dx.doi.org/10.12737/717.
Full textAbstract:
The phenomenon related to a loss of laminar stability of fluid flow in hydrodynamic filter’s working zone with formation of toroidal vortexes is considered. Estimated results related to numerical modeling of liquid’s stationary current in a gap between two coaxial cylinders are presented under various boundary conditions. It is shown that existence of liquid suction from rotating internal cylinder surface leads to stabilization and increase of flow’s stability margin in hydrodynamic filters. The flow stability limits depending on Taylor´s number, rotating cylinder’s angular velocity and liquid suction speed through the cylinder’s surface are defined.
10
Allen, B., T. Pelham, Y. Wu, T. Drysdale, D. Isakov, C. Gamlath, C. J. Stevens, G. Hilton, M. A. Beach, and P. S. Grant. "Experimental evaluation of 3D printed spiral phase plates for enabling an orbital angular momentum multiplexed radio system." Royal Society Open Science 6, no. 12 (December 2019): 191419. http://dx.doi.org/10.1098/rsos.191419.
Full textAbstract:
This paper evaluates the performance of three-dimensionally (3D) printed spiral phase plates (SPPs) for enabling an orbital angular momentum (OAM) multiplexed radio system. The design and realization of the SPPs by means of additive manufacturing exploiting a high-permittivity material is described. Modes 1 and 2 SPPs are then evaluated at 15 GHz in terms of 3D complex radiation pattern, mode purity and beam collimation by means of a 3D printed dielectric lens. The results with the lens yield a crosstalk of −8 dB for between modes 1 and −1, and −11.4 dB for between modes 2 and −2. We suggest a mode multiplexer architecture that is expected to further reduce the crosstalk for each mode. An additional loss of 4.2 dB is incurred with the SPPs inserted into the communication link, which is undesirable for obtaining reliable LTE-based communications. Thus, we suggest: using lower loss materials, seeking ways to reduce material interface reflections or alternative ways of OAM multiplexing to realize a viable OAM multiplexed radio system.
11
Gwizdała, W., A. Dawid, and Z. Gburski. "Molecular Dynamics Simulation Study of the Liquid Crystal Phase in a Small Mesogene Cluster (5CB)22." Solid State Phenomena 140 (October 2008): 89–96. http://dx.doi.org/10.4028/www.scientific.net/ssp.140.89.
Full textAbstract:
The molecular dynamics (MD) technique was used to investigate the nano droplet composed of twenty mesogene molecules 4-cyano-4-n-pentylbiphenyl (5CB). The 5CB molecules were treated as rigid bodies, the intermolecular interaction was taken to be the full site-site pairwise additive Lennard-Jones (LJ) potential plus a Coulomb interaction. The radial distribution functions in the temperature range from 150 to 400 K, were calculated as well as the linear and angular velocity autocorrelation functions. In addition the total dipole moment autocorrelation function and dielectric loss of (5CB)22 mesogene cluster were calculated and the liquid crystal ordering in the nanoscale system was studied up to its vaporization temperature.
12
Xiang, Xiaoyu, Yang Cheng, Jianhang Chen, Qian Lin, and Jan Allebach. "Semi-supervised Multi-task Network For Image Aesthetic Assessment." Electronic Imaging 2020, no. 8 (January 26, 2020): 188–1. http://dx.doi.org/10.2352/issn.2470-1173.2020.8.imawm-188.
Full textAbstract:
Image aesthetic assessment has always been regarded as a challenging task because of the variability of subjective preference. Besides, the assessment of a photo is also related to its style, semantic content, etc. Conventionally, the estimations of aesthetic score and style for an image are treated as separate problems. In this paper, we explore the inter-relatedness between the aesthetics and image style, and design a neural network that can jointly categorize image by styles and give an aesthetic score distribution. To this end, we propose a multi-task network (MTNet) with an aesthetic column serving as a score predictor and a style column serving as a style classifier. The angular-softmax loss is applied in training primary style classifiers to maximize the margin among classes in single-label training data; the semi-supervised method is applied to improve the network’s generalization ability iteratively. We combine the regression loss and classification loss in training aesthetic score. Experiments on the AVA dataset show the superiority of our network in both image attributes classification and aesthetic ranking tasks.
13
Kharalgin, S. V., and M. I. Voytovich. "Investigation of the dielectric characteristics of materials manufactured using additive technologies." Russian Technological Journal 9, no. 2 (April 26, 2021): 57–65. http://dx.doi.org/10.32362/2500-316x-2021-9-2-57-65.
Full textAbstract:
Based on the existing methods of measuring the dielectric characteristics of materials, the most optimal method for performing calculations in the electrodynamic computer-aided design system is selected by the finite integration method. Based on the calculated values of the scattering matrix, the permittivity and the tangent of the dielectric loss angle of the printed polymer samples are calculated according to a given algorithm. When evaluating the accuracy of the calculation of the dielectric characteristics, validation was performed for a sample with the specified characteristics. For a sample printed using the technology of fused filament fabrication, the influence of the filling parameters on the dielectric characteristics of the printed model in the X-band of wavelengths was estimated. The description of the model implemented in the computer-aided design system is given. By processing the simulation results, approximating dependences for the permittivity and losses on the degree of filling with the dielectric are obtained. It follows from the calculated angular diagrams that the decrease in the degree of filling of the dielectric directly affects the degree of anisotropy of the polymer obtained during printing in the plane of the extruded layers. This also increases the depth of the extremes observed at angles of 0°, 90° and 180°. The presence of these extremes is directly related to the fact that the force lines of the main wave type in the waveguide are located perpendicular to the wide wall and in a situation where the volumes of air gaps between the cylinders are parallel to the force lines of tension, there is a general decrease in the dielectric constant. For a printed sample consisting of two layers of crossed cylinders, the air volumes are parallel to the lines of force with a period of ninety degrees, which is confirmed by the results obtained. An increase in the depth of the extremes with a decrease in the degree of filling is associated with a corresponding increase in the air space between the cylinders in the layer of the printed polymer.
14
Korus, Jarosław, Rafał Ziobro, Teresa Witczak, Kamila Kapusniak (Jochym), and Lesław Juszczak. "Effect of Octenyl Succinic Anhydride (OSA) Modified Starches on the Rheological Properties of Dough and Characteristic of the Gluten-Free Bread." Molecules 26, no. 8 (April 11, 2021): 2197. http://dx.doi.org/10.3390/molecules26082197.
Full textAbstract:
The study focused on the influence of starch modified by octenyl succinic anhydride (OSA) on the rheological and thermal properties of gluten-free dough containing corn and potato starch with the addition of pectin and guar gum as structure-forming substances. The starch blend used in the original dough recipe was partially (5% to 15%) replaced with OSA starch. The rheological properties of dough samples were determined, and the properties of the resulting bread were analyzed. It was found that the dough samples behaved as weak gels, and the values of storage and loss moduli (G′ and G″, respectively) significantly depended on angular frequency. Various shares of OSA starch in recipes modified dough in different ways, causing changes in its rheological characteristics. The introduction of OSA starch preparations resulted in changes in the bread volume and physical characteristics of the crumb. All the applied preparations caused an increase in bread porosity and the number of pores larger than 5 mm, and there was a parallel decrease in pore density. The presence of OSA starch preparations modified bread texture depending on the amount and type of the applied preparation. The introduction of OSA starches in gluten-free bread formulation caused a significant drop in the enthalpy of retrograded amylopectin decomposition, indicating a beneficial influence of such type of additive on staling retardation in gluten-free bread.
15
Kim, Jeong-Kyun, Myung-Nam Bae, Kang Bok Lee, and Sang Gi Hong. "Identification of Patients with Sarcopenia Using Gait Parameters Based on Inertial Sensors." Sensors 21, no. 5 (March 4, 2021): 1786. http://dx.doi.org/10.3390/s21051786.
Full textAbstract:
Sarcopenia can cause various senile diseases and is a major factor associated with the quality of life in old age. To diagnose, assess, and monitor muscle loss in daily life, 10 sarcopenia and 10 normal subjects were selected using lean mass index and grip strength, and their gait signals obtained from inertial sensor-based gait devices were analyzed. Given that the inertial sensor can measure the acceleration and angular velocity, it is highly useful in the kinematic analysis of walking. This study detected spatial-temporal parameters used in clinical practice and descriptive statistical parameters for all seven gait phases for detailed analyses. To increase the accuracy of sarcopenia identification, we used Shapley Additive explanations to select important parameters that facilitated high classification accuracy. Support vector machines (SVM), random forest, and multilayer perceptron are classification methods that require traditional feature extraction, whereas deep learning methods use raw data as input to identify sarcopenia. As a result, the input that used the descriptive statistical parameters for the seven gait phases obtained higher accuracy. The knowledge-based gait parameter detection was more accurate in identifying sarcopenia than automatic feature selection using deep learning. The highest accuracy of 95% was achieved using an SVM model with 20 descriptive statistical parameters. Our results indicate that sarcopenia can be monitored with a wearable device in daily life.
16
Żmudzki, J., M. Burzyński, G. Chladek, and C. Krawczyk. "CAD/CAM silicone auricular prosthesis with thermoformed stiffening insert." Archives of Materials Science and Engineering 1, no. 83 (January 2, 2017): 30–35. http://dx.doi.org/10.5604/01.3001.0009.7539.
Full textAbstract:
Purpose: Epitheses (facial prostheses) for large facial tissue defects manufacturedfrom silicones exhibit unsatisfactory rigidity and its stiffening is required, which createstechnological problems. Moreover, facial epitheses have to be replaced in a relatively shortperiod of use which creates a significant costs, often impossible to realize. The hypothesisof the study was that with use of additive manufacturing is possible to obtain the reusableform for thermoforming of the stiffening insert of auricular prosthesis and the mould whichallows multiple casting of silicone prosthesis with the insert.Design/methodology/approach: Manufacturing of the epithesis consisted of designingand manufacturing. In the first step, the auricular prosthesis and the stiffening insert weredesigned with use of engineering CAD software. In this first computer step, the split form forvacuum thermoforming of the stiffening insert and the split mould for casting of the siliconeear were designed. In the second step, additive printing was used for manufacturing thesplit and reusable model for vacuum thermoforming of the stiffening insert and the splitform of ear. In the third step, stiffening insert was made of thermoformed polyurethane sheetof 0.1 mm thick (Biolon, Dreve), where dental thermoforming machine (Drufosmart, Dreve)was used. In the fourth step, the stiffening insert was located in the mould and the ear wascasting of silicone.Findings: CAD/CAM of epitheses with stiffening insert for large tissues defect/loss wasproposed, where in case of re-producing, it required only thermoforming of insert andcasting silicone with use of the reusable models. Dental technician, in case of damage orloss of a forms, is not much involved in their creation.Research limitations/implications: Bond strength test between stiffening insert vs. softsilicone and manufacturing tolerance of epitheses have not been investigated.Practical implications: Method of casting in a negative form, despite the more timeconsumingwhen comparing with epithesis direct-printing, allows introducing a stiffeninginsert and performing a manual adjustment of margin shape and thickness. Method ofnegative form allows the use of a commercially available medical silicone without the needfor medical tests of a new printed materials.
17
Gressel, Jonathan. "Addressing Real Weed Science Needs with Innovations." Weed Technology 6, no. 3 (September 1992): 509–25. http://dx.doi.org/10.1017/s0890037x00035739.
Full textAbstract:
Weed science is responsible for reducing weed competition with crops, within constraints, allowing food production for a growing world population. We have been successful, but with an increasing reliance on the fantastic arsenal of herbicides. Heavy herbicide usage and the concomitant loss of control due to changing weed species spectra, as well as evolved herbicide resistance, along with cancellations of registrations, limit chemical options. Old “innovations” such as well-conceived rotations and various mechanical procedures partly alleviate chemical dependency. We must try innovative, meaningful herbicide mixtures, whether synergistic or additive, mixtures of crop varieties, more competitive varieties, including allelopathic varieties, which could all help to lower chemical dependency. Three cases where innovative genetic-engineering coupled with chemistry seem to be needed are: (a) for the control of parasitic higher plant weeds such as broomrapes (Orobanchespp.), dodders (Cuscutaspp.), witchweeds (Strigaspp.) that cannot be controlled by mechanical or selective chemical means with a sufficient margin of user error; (b) for the control of weeds in wheat (Triticumspp.) that are evolving multiple and cross resistances toallwheat-selective herbicides; and (c) to replace the majors-triazine and chloroacetamide herbicides in corn (Zea maysL.) that are under attack and where many local restrictions and cancellations have been imposed. We need better replacements than the resistance-prone inhibitors of acetolactate-synthase and acetyl-CoA-carboxylase. Much of the long-term innovative planning and research must come from the public sector as no one else seems willing to fill the needs.
18
Dai, Jennifer B., Adam Y. Li, Syed F. Haider, Raechel Tomaselli, Alex Gometz, Stanislaw Sobotka, Alexander F. Post, et al. "Effects of Game Characteristics and Player Positions on Concussion Incidence and Severity in Professional Football." Orthopaedic Journal of Sports Medicine 6, no. 12 (December 1, 2018): 232596711881544. http://dx.doi.org/10.1177/2325967118815448.
Full textAbstract:
Background: Increasing efforts have been made to reduce the incidence and severity of concussion in high-contact sports. Despite these efforts, a relative lack of knowledge is available regarding modulating factors affecting concussion injury. Purpose: To analyze the potential influence of game characteristics and outcomes on concussion incidence and severity in professional football. Study Design: Descriptive epidemiology study. Methods: PBS Frontline Concussion Watch was used to collect concussion injury data from regular-season games of 32 National Football League (NFL) teams from 2012 to 2015. Game characteristic variables such as rushing and passing attempts, turnovers, and margin of victory were collected from ESPN. Analysis included descriptive statistics, analysis of variance, t tests, and correlation tests. Results: Away teams demonstrated a significantly greater concussion incidence per game than home teams. Losing teams had a significantly greater concussion incidence per game than winning teams. Being both the away team and the losing team appeared to have an additive effect. The home-versus-away and win-versus-loss effects were significant for offensive but not defensive positions. Within individual positions, significantly greater concussion incidence was associated with tight ends, running backs, wide receivers, and cornerbacks. When running versus passing positions were compared, passing positions (wide receiver, tight end, cornerback, safety) had significantly greater concussion incidence. A total of 626 games were missed as a result of reported concussions. Away teams had significantly more games missed due to concussion when they lost. Play time did not significantly differ before or after concussion injury. Other game characteristic variables did not significantly affect concussion frequency or intensity. Conclusion: Position, game location, and game outcome affect concussion incidence for professional football players. In a subset of analyses, the number of games missed aligned with concussion incidence, but this appeared to be an imperfect measure. These findings highlight new factors that may modulate concussion incidence and merit further study on how they may influence concussion evaluation.
19
Blomquist, C. L., and C. Y. Warfield. "First Report of Ramularia didyma Causing a Leaf Spot on Ranunculus (Ranunculus asiaticus) Hybrids in California." Plant Disease 95, no. 7 (July 2011): 872. http://dx.doi.org/10.1094/pdis-12-10-0892.
Full textAbstract:
Ranunculus or Persian buttercups (Ranunculus asiaticus) are colorful, cool-season perennials or annuals grown as landscape bedding plants and for field-grown bulb and cut flower production in mild winter climates. In March of 2008, tan-to-brown lesions were observed on the foliage of containerized ranunculus growing in a greenhouse at a production and retail nursery in coastal San Mateo County, CA. Approximately 15% of the approximately 150 cv. Bloomingdale mixed shade ranunculus plants had leaf spot symptoms. Symptomatic plants were generally clustered together on two benches, with a double-flowered purple cultivar with picotee markings having the highest disease incidence and severity. Angular to round, necrotic lesions ranged between 0.25 and 1.0 cm in diameter and were often surrounded by chlorotic tissue. Whitish sporulation was observed on abaxial and adaxial sides of the larger lesions. Conidiophores were hyaline, aseptate, measured 45 to 75 × 3 to 5 μm (56 × 3.7 μm average), and were produced in fascicles on the leaf surface. One-celled fusiform to cylindrical conidia measured 22.5 to 30.0 × 9.0 to 15 μm (25.4 × 11.9 μm average) and two-celled conidia measured 26.0 to 47.5 × 10 to 14 μm (35.3 × 12.1 μm average). Most conidia were hyaline, although a few were pale brown. Morphologically, the fungus matched the description of Ramularia didyma Unger (1). Small (3-mm2) pieces were taken from the margin of the lesion, disinfested with 0.6% sodium hypochlorite for 2 min, and placed at room temperature on carrot piece agar. Colonies were white and grew slowly to approximately 3.5 cm in 25 days. No conidia were produced in culture. To conduct pathogenicity tests, the foliage of nine mixed-color ranunculus plants grown from tubers for 7 weeks at 16°C in a growth chamber were sprayed with an aqueous suspension of mycelia. Inoculum was produced by grinding five, 3.5-cm colonies on carrot piece agar in 120 ml of water with a mortar and pestle. The foliage of an equal number of plants was sprayed with water. Plants were incubated in a dew chamber at 20°C for 48 h in the dark and then moved into a growth chamber with a 12-h photoperiod where relative humidity was maintained at ~95% by placing plants over a tray of water and covering each plant group with a plastic tent. Small, angular spots developed on approximately half of the inoculated plants 20 days later when the fungus was reisolated from the lesions. No symptoms were observed on the noninoculated control plants. Sporulation was observed on diseased lesions following misting of plants and incubation in sealed plastic bags for an additional week. Pathogenicity tests were repeated with mycelia with similar results. Sequence of the internal transcribed spacer regions of the rDNA was deposited into GenBank (Accession No. HQ442297). The sequence matched R. coleosporii and R. carthami with 96% identity. R. didyma has been reported to cause a leaf spot on Ranunculus spp. in the United States (Delaware, Iowa, Indiana, Maryland, New York, and Vermont). To our knowledge, this is the first confirmed report of R. didyma on Ranunculus asiaticus in California. Introduction of this pathogen into commercial production fields could cause significant economic loss. The closest large-scale production fields are located approximately 430 miles south of San Mateo County near Carlsbad, CA where more than 50 acres of ranunculus are grown for cut flower and tuber production. Reference: (1) J. B. Ellis and B. M. Everhart. J. Mycol. 1:79, 1885.
20
Jayakumar, V., G. Kannamma Usha Rani, N. Amaresan, and S. Rajalakshmi. "First Report of Anthracnose Disease of Black Pepper (Piper nigrum) Caused by an Unknown Species of Colletotrichum." Plant Disease 93, no. 2 (February 2009): 199. http://dx.doi.org/10.1094/pdis-93-2-0199a.
Full textAbstract:
Black pepper is cultivated in Andaman and Nicobar Islands, India as a spice crop. During a survey performed in June of 2007 in South Andaman, two kinds of leaf anthracnose symptoms were observed. The classic symptom, angular to irregular or circular brownish lesions with a chlorotic halo and pinhead size acervuli on the leaves, was consistent with the disease previously reported on pepper caused by Colletotrichum gloeosporioides (2). This symptom also caused splitting and the production of hollow berries. The new symptom was novel, with leaves initially exhibiting pale green or yellowish green lesions. As the disease progressed, lesion margins became brown to black with slightly raised areas containing numerous acervuli. Unlike the classic symptom, the new symptom was characterized as leaf lesions that rarely caused defoliation; berries showed no lesions. The foliar disease incidence was up to 15%, but direct economic loss of berries was not noticed. This new symptom was similar to symptoms caused by C. dracaenophilum, a species identified as a pathogen on lucky bamboo (Dracaena sanderiana) (1). A fungus was isolated in pure culture from the pale green lesions and cultured on potato dextrose agar (PDA). At room temperature (28 ± 2°C), fungal growth on PDA was slow and mycelium appeared whitish at the margin with pale pink centers. A pinkish color was observed on the reverse side of the plate, reflecting profuse sporulation. The conidia were hyaline, broadly clavate to cylindrical, and measured 12.5 to 15 × 5 to 7.5 μm (average 14 × 7.5 μm). The internal transcribed spacer (ITS) region of the fungal DNA was amplified, sequenced, and submitted to NCBI GenBank (Accession No. EU744584). The specimen was deposited in the MTCC of IMTECH, Chandigarh, India (Accession No. MTCC9344). Pathogenicity was tested in five replications on 15- to 20-day-old pepper plants and repeated twice. A 1-ml conidial suspension (108 spores/ml) of the fungus was brushed on two intact leaves of each pepper plant and incubated for 2 weeks in a glasshouse at 28°C and 70% relative humidity with natural daylight conditions. Plants brushed with sterile water served as control. Similar pale green symptoms were observed only on treated leaves and the same organism was reisolated from lesions. BLAST searches of the GenBank using the ITS sequence revealed that this fungus was a member of the genus Colletotrichum, but a species level identification could not be made with these data. The fungus was most similar in sequence to unnamed endophytic strains of Colletotrichum (96% sequence identity) and phytopathogenic isolates of C. dracaenophilum (93% sequence similarity). Although the symptomatology and sequence data were most closely matched with those documented for C. dracaenophilum (1), the morphological and cultural characteristics of the black pepper anthracnose fungus differed from C. dracaenophilum and other known species of Colletotrichum (3). Together these morphological and molecular data suggest that this form of anthracnose disease on black pepper may be caused by a novel, undescribed species of Colletotrichum. Further investigations will be required to characterize this organism to the species level. References: (1) S. G. Bobev et al. Plant Dis. 92:173, 2008. (2) P. Santha Kumari and A. Sanker. J. Mycol. Plant Pathol. 33:329, 2003. (3) B. C. Sutton. In: Colletotrichum. Biology, Pathology and Control. CAB International, Wallingford, 1992.
21
Deng, Jiankang, Jia Guo, Jing Yang, Niannan Xue, Irene Cotsia, and Stefanos P. Zafeiriou. "ArcFace: Additive Angular Margin Loss for Deep Face Recognition." IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021, 1. http://dx.doi.org/10.1109/tpami.2021.3087709.
Full text
22
Jiao, Jichao, Weilun Liu, Yaokai Mo, Jian Jiao, Zhongliang Deng, and Xinping Chen. "Dyn-arcFace: dynamic additive angular margin loss for deep face recognition." Multimedia Tools and Applications, April 23, 2021. http://dx.doi.org/10.1007/s11042-021-10865-5.
Full text
23
Yang, Hang, Xin-Rong Hu, Ling Sun, Dian Hong, Ying-Yi Zheng, Ying Xin, Hui Liu, et al. "Automated Facial Recognition for Noonan Syndrome Using Novel Deep Convolutional Neural Network With Additive Angular Margin Loss." Frontiers in Genetics 12 (June 7, 2021). http://dx.doi.org/10.3389/fgene.2021.669841.
Full textAbstract:
BackgroundNoonan syndrome (NS), a genetically heterogeneous disorder, presents with hypertelorism, ptosis, dysplastic pulmonary valve stenosis, hypertrophic cardiomyopathy, and small stature. Early detection and assessment of NS are crucial to formulating an individualized treatment protocol. However, the diagnostic rate of pediatricians and pediatric cardiologists is limited. To overcome this challenge, we propose an automated facial recognition model to identify NS using a novel deep convolutional neural network (DCNN) with a loss function called additive angular margin loss (ArcFace).MethodsThe proposed automated facial recognition models were trained on dataset that included 127 NS patients, 163 healthy children, and 130 children with several other dysmorphic syndromes. The photo dataset contained only one frontal face image from each participant. A novel DCNN framework with ArcFace loss function (DCNN-Arcface model) was constructed. Two traditional machine learning models and a DCNN model with cross-entropy loss function (DCNN-CE model) were also constructed. Transfer learning and data augmentation were applied in the training process. The identification performance of facial recognition models was assessed by five-fold cross-validation. Comparison of the DCNN-Arcface model to two traditional machine learning models, the DCNN-CE model, and six physicians were performed.ResultsAt distinguishing NS patients from healthy children, the DCNN-Arcface model achieved an accuracy of 0.9201 ± 0.0138 and an area under the receiver operator characteristic curve (AUC) of 0.9797 ± 0.0055. At distinguishing NS patients from children with several other genetic syndromes, it achieved an accuracy of 0.8171 ± 0.0074 and an AUC of 0.9274 ± 0.0062. In both cases, the DCNN-Arcface model outperformed the two traditional machine learning models, the DCNN-CE model, and six physicians.ConclusionThis study shows that the proposed DCNN-Arcface model is a promising way to screen NS patients and can improve the NS diagnosis rate.
24
Ma, Yuandong, Shouyu Sun, Fengjiao Wu, Yunfan Yang, Xin Yang, Bin Xu, and Zijiang Luo. "Additive margin cosine loss for image registration." Visual Computer, March 18, 2021. http://dx.doi.org/10.1007/s00371-021-02105-6.
Full text
25
Houghton, Tim, and Ivor Day. "Stability Enhancement by Casing Grooves: The Importance of Stall Inception Mechanism and Solidity." Journal of Turbomachinery 134, no. 2 (June 21, 2011). http://dx.doi.org/10.1115/1.4002986.
Full textAbstract:
This paper concerns the optimization of casing grooves and the important influence of stall inception mechanism on groove performance. Installing casing grooves is a well known technique for improving the stable operating range of a compressor, but the wide-spread use of grooves is restricted by the loss of efficiency and flow capacity. In this paper, laboratory tests are used to examine the conditions under which casing treatment can be used to greatest effect. The use of a single casing groove was investigated in a recently published companion paper. The current work extends this to multiple-groove treatments and considers their performance in relation to stall inception mechanisms. Here it is shown that the stall margin gain from multiple grooves is less than the sum of the gains if the grooves were used individually. By contrast, the loss of efficiency is additive as the number of grooves increases. It is then shown that casing grooves give the greatest stall margin improvement when used in a compressor, that exhibits spike-type stall inception, while modal activity before stall can dramatically reduce the effectiveness of the grooves. This finding highlights the importance of being able to predict which stall inception mechanism might occur in a given compressor before and after grooves are added. Some published prediction techniques are therefore examined, but found wanting. Lastly, it is shown that casing grooves can, in some cases, be used to remove rotor blades and produce a more efficient, stable, and light-weight rotor.
26
Xiao, Bing, Qinglei Hu, and Michael I. Friswell. "Robust Fault Tolerant Control for Spacecraft Attitude Stabilization Under Actuator Faults and Bounded Disturbance." Journal of Dynamic Systems, Measurement, and Control 133, no. 5 (August 1, 2011). http://dx.doi.org/10.1115/1.4004061.
Full textAbstract:
This paper investigates the design of spacecraft attitude stabilization controllers that are robust against actuator faults and external disturbances. A nominal controller is developed initially, using the adaptive backstepping technique, to stabilize asymptotically the spacecraft attitude when the actuators are fault-free. Additive faults and the partial loss of actuator effectiveness are considered simultaneously and an auxiliary controller is designed in addition to the nominal controller to compensate for the system faults. This auxiliary controller does not use any fault detection and isolation mechanism to detect, separate, and identify the actuator faults online. The attitude orientation and angular velocity of the closed-loop system asymptotically converge to zero despite actuator faults providing the nominal attitude system is asymptotically stable. Numerical simulation results are presented that demonstrate the closed-loop performance benefits of the proposed control law and illustrate its robustness to external disturbances and actuator faults.
27
Praniewicz, Maxwell, Thomas Kurfess, and Christopher Saldana. "An Adaptive Geometry Transformation and Repair Method for Hybrid Manufacturing." Journal of Manufacturing Science and Engineering 141, no. 1 (October 19, 2018). http://dx.doi.org/10.1115/1.4041570.
Full textAbstract:
Hybrid manufacturing has become particularly attractive for refurbishing of high-value freeform components. Components may experience unique geometric distortions and/or wear-driven material loss in service, which require the use of part-specific, adaptive repair strategies. The current work presents an integrated adaptive geometry transformation method for additive/subtractive hybrid manufacturing based on rigid and nonrigid registrations of parent region material and geometric interpolation of the repair region material. In this approach, rigid registration of nominal part geometry to actual part geometry is accomplished using iterative alignment of profiles in the parent material. Nonrigid registration is used to morph nominal part geometry to actual part geometry by transformation of the profile mean line. Adaptive additive and subtractive tool paths are then used to add material based on constant stock margin requirements, as well as to produce blend repairs with smooth transition between parent and repair regions. A range of part deformation conditions due to profile twist and length changes are evaluated for the case of a compressor blade/airfoil geometry. Accuracy of the resulting adaptive geometry transformation method were quantified by (1) surface comparisons of actual and transformed nominal geometry and (2) blend region surface accuracy. Performance of the adaptive repair strategy relative to a naïve strategy is evaluated by the consideration of material efficiency and process cycle time. It is shown that the adaptive repair strategy resulted in an increase in material efficiency by 42.2% and a decrease in process time by 17.8%, depending on the initial deformation imposed on the part geometry.