Статті в журналах з теми "Generative competitive neural network"
Щоб переглянути інші типи публікацій з цієї теми, перейдіть за посиланням: Generative competitive neural network.
Оформте джерело за APA, MLA, Chicago, Harvard та іншими стилями
Ознайомтеся з топ-50 статей у журналах для дослідження на тему "Generative competitive neural network".
Біля кожної праці в переліку літератури доступна кнопка «Додати до бібліографії». Скористайтеся нею – і ми автоматично оформимо бібліографічне посилання на обрану працю в потрібному вам стилі цитування: APA, MLA, «Гарвард», «Чикаго», «Ванкувер» тощо.
Також ви можете завантажити повний текст наукової публікації у форматі «.pdf» та прочитати онлайн анотацію до роботи, якщо відповідні параметри наявні в метаданих.
Переглядайте статті в журналах для різних дисциплін та оформлюйте правильно вашу бібліографію.
1
Kuznetsov, A. V., and M. V. Gashnikov. "Remote sensing data retouching based on image inpainting algorithms in the forgery generation problem." Computer Optics 44, no. 5 (October 2020): 763–71. http://dx.doi.org/10.18287/2412-6179-co-721.
Повний текст джерелаАнотація:
We investigate image retouching algorithms for generating forgery Earth remote sensing data. We provide an overview of existing neural network solutions in the field of generation and inpainting of remote sensing images. To retouch Earth remote sensing data, we use imageinpainting algorithms based on convolutional neural networks and generative-adversarial neural networks. We pay special attention to a generative neural network with a separate contour prediction block that includes two series-connected generative-adversarial subnets. The first subnet inpaints contours of the image within the retouched area. The second subnet uses the inpainted contours to generate the resulting retouch area. As a basis for comparison, we use exemplar-based algorithms of image inpainting. We carry out computational experiments to study the effectiveness of these algorithms when retouching natural data of remote sensing of various types. We perform a comparative analysis of the quality of the algorithms considered, depending on the type, shape and size of the retouched objects and areas. We give qualitative and quantitative characteristics of the efficiency of the studied image inpainting algorithms when retouching Earth remote sensing data. We experimentally prove the advantage of generative-competitive neural networks in the construction of forgery remote sensing data.
2
Tsibulis, Dmitry E., Andrey N. Ragozin, Stanislav N. Darovskikh, and Askar Z. Kulganatov. "Study of nonlinear digital filtering of signals using generative competitive neural network." Bulletin of the South Ural State University. Ser. Computer Technologies, Automatic Control & Radioelectronics 22, no. 2 (April 2022): 158–67. http://dx.doi.org/10.14529/ctcr220215.
Повний текст джерелаАнотація:
The article presents the results of the study, as well as the structural schemes and parameters of the components of the generative-adversarial neural network. Graphical images of the results of filtering radio signals are given. Conclusions are drawn about the possibilities of using these neural networks. The purpose of the study. Substantiation of the possibilities of using generative-sensory artificial neural networks to solve problems of digital processing of radio signals. Materials and methods. To evaluate the results of digital filtering of noisy signals, the method of mathematical modeling in the Matlab environment was used. As test signals, the following were taken: a sine wave, a signal in the form of a sum of sinusoids, a model of a real radio-technical information signal. White Gaussian noise is used as the noise component. Also, filtering of the signal is carried out, in which there is no fragment of a certain length. A training sample was generated for the neural network of the generator, consisting of noisy test signals. A training sample of the discriminator neural network was also generated, consisting of test signals that do not contain noise. Results. Based on the simulation, it is concluded that the generative-adversarial neural network successfully solves the problems of isolating a useful signal in a mixture of it with noise of various physical nature. Such a neural network structure is also able to restore a useful signal if any part of it is missing as a result of external interference. Conclusion. The existing methods of digital filtering of radio signals require certain labor and time costs associated with the calculation of digital filters. Also, when designing high-order filters, it becomes difficult to calculate these filters. The idea of using a neural network in filtering tasks makes it possible to significantly reduce the filter design time, thus simplifying the process of its implementation. A neural network, which is a self-learning system, can find solutions that are inaccessible to conventional digital filtering algorithms. The results of this work can find their application in the field of digital signal processing and in the development of software-configurable radio.
3
Marton, Sascha, Stefan Lüdtke, and Christian Bartelt. "Explanations for Neural Networks by Neural Networks." Applied Sciences 12, no. 3 (January 18, 2022): 980. http://dx.doi.org/10.3390/app12030980.
Повний текст джерелаАнотація:
Understanding the function learned by a neural network is crucial in many domains, e.g., to detect a model’s adaption to concept drift in online learning. Existing global surrogate model approaches generate explanations by maximizing the fidelity between the neural network and a surrogate model on a sample-basis, which can be very time-consuming. Therefore, these approaches are not applicable in scenarios where timely or frequent explanations are required. In this paper, we introduce a real-time approach for generating a symbolic representation of the function learned by a neural network. Our idea is to generate explanations via another neural network (called the Interpretation Network, or I-Net), which maps network parameters to a symbolic representation of the network function. We show that the training of an I-Net for a family of functions can be performed up-front and subsequent generation of an explanation only requires querying the I-Net once, which is computationally very efficient and does not require training data. We empirically evaluate our approach for the case of low-order polynomials as explanations, and show that it achieves competitive results for various data and function complexities. To the best of our knowledge, this is the first approach that attempts to learn mapping from neural networks to symbolic representations.
4
Shin, Wonsup, Seok-Jun Bu, and Sung-Bae Cho. "3D-Convolutional Neural Network with Generative Adversarial Network and Autoencoder for Robust Anomaly Detection in Video Surveillance." International Journal of Neural Systems 30, no. 06 (May 28, 2020): 2050034. http://dx.doi.org/10.1142/s0129065720500343.
Повний текст джерелаАнотація:
As the surveillance devices proliferate, various machine learning approaches for video anomaly detection have been attempted. We propose a hybrid deep learning model composed of a video feature extractor trained by generative adversarial network with deficient anomaly data and an anomaly detector boosted by transferring the extractor. Experiments with UCSD pedestrian dataset show that it achieves 94.4% recall and 86.4% precision, which is the competitive performance in video anomaly detection.
5
Forster, Dennis, Abdul-Saboor Sheikh, and Jörg Lücke. "Neural Simpletrons: Learning in the Limit of Few Labels with Directed Generative Networks." Neural Computation 30, no. 8 (August 2018): 2113–74. http://dx.doi.org/10.1162/neco_a_01100.
Повний текст джерелаАнотація:
We explore classifier training for data sets with very few labels. We investigate this task using a neural network for nonnegative data. The network is derived from a hierarchical normalized Poisson mixture model with one observed and two hidden layers. With the single objective of likelihood optimization, both labeled and unlabeled data are naturally incorporated into learning. The neural activation and learning equations resulting from our derivation are concise and local. As a consequence, the network can be scaled using standard deep learning tools for parallelized GPU implementation. Using standard benchmarks for nonnegative data, such as text document representations, MNIST, and NIST SD19, we study the classification performance when very few labels are used for training. In different settings, the network's performance is compared to standard and recently suggested semisupervised classifiers. While other recent approaches are more competitive for many labels or fully labeled data sets, we find that the network studied here can be applied to numbers of few labels where no other system has been reported to operate so far.
6
Wang, Zheng, and Qingbiao Wu. "An Integrated Deep Generative Model for Text Classification and Generation." Mathematical Problems in Engineering 2018 (August 19, 2018): 1–8. http://dx.doi.org/10.1155/2018/7529286.
Повний текст джерелаАнотація:
Text classification and generation are two important tasks in the field of natural language processing. In this paper, we deal with both tasks via Variational Autoencoder, which is a powerful deep generative model. The self-attention mechanism is introduced to the encoder. The modified encoder extracts the global feature of the input text to produce the hidden code, and we train a neural network classifier based on the hidden code to perform the classification. On the other hand, the label of the text is fed into the decoder explicitly to enhance the categorization information, which could help with text generation. The experiments have shown that our model could achieve competitive classification results and the generated text is realistic. Thus the proposed integrated deep generative model could be an alternative for both tasks.
7
Huang, Wenlong, Brian Lai, Weijian Xu, and Zhuowen Tu. "3D Volumetric Modeling with Introspective Neural Networks." Proceedings of the AAAI Conference on Artificial Intelligence 33 (July 17, 2019): 8481–88. http://dx.doi.org/10.1609/aaai.v33i01.33018481.
Повний текст джерелаАнотація:
In this paper, we study the 3D volumetric modeling problem by adopting the Wasserstein introspective neural networks method (WINN) that was previously applied to 2D static images. We name our algorithm 3DWINN which enjoys the same properties as WINN in the 2D case: being simultaneously generative and discriminative. Compared to the existing 3D volumetric modeling approaches, 3DWINN demonstrates competitive results on several benchmarks in both the generation and the classification tasks. In addition to the standard inception score, the Frechet Inception Distance (FID) metric is´ also adopted to measure the quality of 3D volumetric generations. In addition, we study adversarial attacks for volumetric data and demonstrate the robustness of 3DWINN against adversarial examples while achieving appealing results in both classification and generation within a single model. 3DWINN is a general framework and it can be applied to the emerging tasks for 3D object and scene modeling.1
8
V.M., Sineglazov, and Chumachenko O.I. "Structural-parametric synthesis of deep learning neural networks." Artificial Intelligence 25, no. 4 (December 25, 2020): 42–51. http://dx.doi.org/10.15407/jai2020.04.042.
Повний текст джерелаАнотація:
The structural-parametric synthesis of neural networks of deep learning, in particular convolutional neural networks used in image processing, is considered. The classification of modern architectures of convolutional neural networks is given. It is shown that almost every convolutional neural network, depending on its topology, has unique blocks that determine its essential features (for example, Squeeze and Excitation Block, Convolutional Block of Attention Module (Channel attention module, Spatial attention module), Residual block, Inception module, ResNeXt block. It is stated the problem of structural-parametric synthesis of convolutional neural networks, for the solution of which it is proposed to use a genetic algorithm. The genetic algorithm is used to effectively overcome a large search space: on the one hand, to generate possible topologies of the convolutional neural network, namely the choice of specific blocks and their locations in the structure of the convolutional neural network, and on the other hand to solve the problem of structural-parametric synthesis of convolutional neural network of selected topology. The most significant parameters of the convolutional neural network are determined. An encoding method is proposed that allows to repre- sent each network structure in the form of a string of fixed length in binary format. After that, several standard genetic operations were identified, i.e. selection, mutation and crossover, which eliminate weak individuals of the previous generation and use them to generate competitive ones. An example of solving this problem is given, a database (ultrasound results) of patients with thyroid disease was used as a training sample.
9
Lee, Byong Kwon. "A Combining AI Algorithm for the Restoration of Damaged Cultural Properties." Webology 19, no. 1 (January 20, 2022): 4384–95. http://dx.doi.org/10.14704/web/v19i1/web19288.
Повний текст джерелаАнотація:
Through the research of numerous researchers’ artificial intelligence imitates human language and visual expression with good performance and imitates human style in voice and picture. This ability although dependent on the data for learning artificial intelligence is more objective and based on numerical data than humans. We applied it to the restoration of cultural assets made in the past through artificial intelligence neural networks and we applied a general CNN a little differently for the purpose of restoration. Cultural properties contain various backgrounds from the era when they were created and for this reason there are many complications and difficulties in restoration. If it is simply regarded as noise and recovered the result is dependent on the learned data. To solve this problem the CNN was separated into full and detailed and the association was learned together and the damaged part was repaired through a generative competition network (GAN) based on this neural network. We trained a neural network that extracts visual features on a Korean "Pagoda" (mostly produced under the influence of Buddhism) and conducted a study to repair the damaged part based on the trained neural network. The features of the tower were extracted through a CNN-based neural network and the damaged part was repaired through a Generative Adversarial Network (GAN) based on the extracted features. It is thought that our research will be actively used for the restoration of cultural assets as well as the restoration of archaeological records in the future.
10
Wang, Zilin, Zhaoxiang Zhang, Limin Dong, and Guodong Xu. "Jitter Detection and Image Restoration Based on Generative Adversarial Networks in Satellite Images." Sensors 21, no. 14 (July 9, 2021): 4693. http://dx.doi.org/10.3390/s21144693.
Повний текст джерелаАнотація:
High-resolution satellite images (HRSIs) obtained from onboard satellite linear array cameras suffer from geometric disturbance in the presence of attitude jitter. Therefore, detection and compensation of satellite attitude jitter are crucial to reduce the geopositioning error and to improve the geometric accuracy of HRSIs. In this work, a generative adversarial network (GAN) architecture is proposed to automatically learn and correct the deformed scene features from a single remote sensing image. In the proposed GAN, a convolutional neural network (CNN) is designed to discriminate the inputs, and another CNN is used to generate so-called fake inputs. To explore the usefulness and effectiveness of a GAN for jitter detection, the proposed GANs are trained on part of the PatternNet dataset and tested on three popular remote sensing datasets, along with a deformed Yaogan-26 satellite image. Several experiments show that the proposed model provides competitive results. The proposed GAN reveals the enormous potential of GAN-based methods for the analysis of attitude jitter from remote sensing images.
11
KOHLMORGEN, JENS, and BENJAMIN BLANKERTZ. "BAYESIAN CLASSIFICATION OF SINGLE-TRIAL EVENT-RELATED POTENTIALS IN EEG." International Journal of Bifurcation and Chaos 14, no. 02 (February 2004): 719–26. http://dx.doi.org/10.1142/s0218127404009429.
Повний текст джерелаАнотація:
We present a systematic and straightforward approach to the problem of single-trial classification of event-related potentials (ERP) in EEG. Instead of using a generic classifier off-the-shelf, like a neural network or support vector machine, our classifier design is guided by prior knowledge about the problem and statistical properties found in the data. In particular, we exploit the well-known fact that event-related drifts in EEG potentials, albeit hard to detect in a single trial, can well be observed if averaged over a sufficiently large number of trials. We propose to use the average signal and its variance as a generative model for each event class and use Bayes' decision rule for the classification of new and unlabeled data. The method is successfully applied to a data set from the NIPS*2001 Brain–Computer Interface post-workshop competition. Our result turned out to be competitive with the best result of the competition.
12
Lobov, Sergey A., Andrey V. Chernyshov, Nadia P. Krilova, Maxim O. Shamshin, and Victor B. Kazantsev. "Competitive Learning in a Spiking Neural Network: Towards an Intelligent Pattern Classifier." Sensors 20, no. 2 (January 16, 2020): 500. http://dx.doi.org/10.3390/s20020500.
Повний текст джерелаАнотація:
One of the modern trends in the design of human–machine interfaces (HMI) is to involve the so called spiking neuron networks (SNNs) in signal processing. The SNNs can be trained by simple and efficient biologically inspired algorithms. In particular, we have shown that sensory neurons in the input layer of SNNs can simultaneously encode the input signal based both on the spiking frequency rate and on varying the latency in generating spikes. In the case of such mixed temporal-rate coding, the SNN should implement learning working properly for both types of coding. Based on this, we investigate how a single neuron can be trained with pure rate and temporal patterns, and then build a universal SNN that is trained using mixed coding. In particular, we study Hebbian and competitive learning in SNN in the context of temporal and rate coding problems. We show that the use of Hebbian learning through pair-based and triplet-based spike timing-dependent plasticity (STDP) rule is accomplishable for temporal coding, but not for rate coding. Synaptic competition inducing depression of poorly used synapses is required to ensure a neural selectivity in the rate coding. This kind of competition can be implemented by the so-called forgetting function that is dependent on neuron activity. We show that coherent use of the triplet-based STDP and synaptic competition with the forgetting function is sufficient for the rate coding. Next, we propose a SNN capable of classifying electromyographical (EMG) patterns using an unsupervised learning procedure. The neuron competition achieved via lateral inhibition ensures the “winner takes all” principle among classifier neurons. The SNN also provides gradual output response dependent on muscular contraction strength. Furthermore, we modify the SNN to implement a supervised learning method based on stimulation of the target classifier neuron synchronously with the network input. In a problem of discrimination of three EMG patterns, the SNN with supervised learning shows median accuracy 99.5% that is close to the result demonstrated by multi-layer perceptron learned by back propagation of an error algorithm.
13
Chadha, Aman, John Britto, and M. Mani Roja. "iSeeBetter: Spatio-temporal video super-resolution using recurrent generative back-projection networks." Computational Visual Media 6, no. 3 (July 20, 2020): 307–17. http://dx.doi.org/10.1007/s41095-020-0175-7.
Повний текст джерелаАнотація:
Abstract Recently, learning-based models have enhanced the performance of single-image super-resolution (SISR). However, applying SISR successively to each video frame leads to a lack of temporal coherency. Convolutional neural networks (CNNs) outperform traditional approaches in terms of image quality metrics such as peak signal to noise ratio (PSNR) and structural similarity (SSIM). On the other hand, generative adversarial networks (GANs) offer a competitive advantage by being able to mitigate the issue of a lack of finer texture details, usually seen with CNNs when super-resolving at large upscaling factors. We present iSeeBetter, a novel GAN-based spatio-temporal approach to video super-resolution (VSR) that renders temporally consistent super-resolution videos. iSeeBetter extracts spatial and temporal information from the current and neighboring frames using the concept of recurrent back-projection networks as its generator. Furthermore, to improve the “naturality” of the super-resolved output while eliminating artifacts seen with traditional algorithms, we utilize the discriminator from super-resolution generative adversarial network. Although mean squared error (MSE) as a primary loss-minimization objective improves PSNR/SSIM, these metrics may not capture fine details in the image resulting in misrepresentation of perceptual quality. To address this, we use a four-fold (MSE, perceptual, adversarial, and total-variation loss function. Our results demonstrate that iSeeBetter offers superior VSR fidelity and surpasses state-of-the-art performance.
14
Gao, Wei, Linjie Zhou, and Lvfang Tao. "A Fast View Synthesis Implementation Method for Light Field Applications." ACM Transactions on Multimedia Computing, Communications, and Applications 17, no. 4 (November 30, 2021): 1–20. http://dx.doi.org/10.1145/3459098.
Повний текст джерелаАнотація:
View synthesis (VS) for light field images is a very time-consuming task due to the great quantity of involved pixels and intensive computations, which may prevent it from the practical three-dimensional real-time systems. In this article, we propose an acceleration approach for deep learning-based light field view synthesis, which can significantly reduce calculations by using compact-resolution (CR) representation and super-resolution (SR) techniques, as well as light-weight neural networks. The proposed architecture has three cascaded neural networks, including a CR network to generate the compact representation for original input views, a VS network to synthesize new views from down-scaled compact views, and a SR network to reconstruct high-quality views with full resolution. All these networks are jointly trained with the integrated losses of CR, VS, and SR networks. Moreover, due to the redundancy of deep neural networks, we use the efficient light-weight strategy to prune filters for simplification and inference acceleration. Experimental results demonstrate that the proposed method can greatly reduce the processing time and become much more computationally efficient with competitive image quality.
15
Du, Bingqian, Zhiyi Huang, and Chuan Wu. "Adversarial Deep Learning for Online Resource Allocation." ACM Transactions on Modeling and Performance Evaluation of Computing Systems 6, no. 4 (December 31, 2021): 1–25. http://dx.doi.org/10.1145/3494526.
Повний текст джерелаАнотація:
Online algorithms are an important branch in algorithm design. Designing online algorithms with a bounded competitive ratio (in terms of worst-case performance) can be hard and usually relies on problem-specific assumptions. Inspired by adversarial training from Generative Adversarial Net and the fact that the competitive ratio of an online algorithm is based on worst-case input, we adopt deep neural networks (NNs) to learn an online algorithm for a resource allocation and pricing problem from scratch, with the goal that the performance gap between offline optimum and the learned online algorithm can be minimized for worst-case input. Specifically, we leverage two NNs as the algorithm and the adversary, respectively, and let them play a zero sum game, with the adversary being responsible for generating worst-case input while the algorithm learns the best strategy based on the input provided by the adversary. To ensure better convergence of the algorithm network (to the desired online algorithm), we propose a novel per-round update method to handle sequential decision making to break complex dependency among different rounds so that update can be done for every possible action instead of only sampled actions. To the best of our knowledge, our work is the first using deep NNs to design an online algorithm from the perspective of worst-case performance guarantee. Empirical studies show that our updating methods ensure convergence to Nash equilibrium and the learned algorithm outperforms state-of-the-art online algorithms under various settings.
16
Usoltsev, Yakov, Balzhit Lodonova, Alexander Shelupanov, Anton Konev, and Evgeny Kostyuchenko. "Adversarial Attacks Impact on the Neural Network Performance and Visual Perception of Data under Attack." Information 13, no. 2 (February 5, 2022): 77. http://dx.doi.org/10.3390/info13020077.
Повний текст джерелаАнотація:
Machine learning algorithms based on neural networks are vulnerable to adversarial attacks. The use of attacks against authentication systems greatly reduces the accuracy of such a system, despite the complexity of generating a competitive example. As part of this study, a white-box adversarial attack on an authentication system was carried out. The basis of the authentication system is a neural network perceptron, trained on a dataset of frequency signatures of sign. For an attack on an atypical dataset, the following results were obtained: with an attack intensity of 25%, the authentication system availability decreases to 50% for a particular user, and with a further increase in the attack intensity, the accuracy decreases to 5%.
17
Yuan, Yijun, Xu Si, and Yue Zheng. "Ground-roll attenuation using generative adversarial networks." GEOPHYSICS 85, no. 4 (June 13, 2020): WA255—WA267. http://dx.doi.org/10.1190/geo2019-0414.1.
Повний текст джерелаАнотація:
Ground roll is a persistent problem in land seismic data. This type of coherent noise often contaminates seismic signals and severely reduces the signal-to-noise ratio of seismic data. A variety of methods for addressing ground-roll attenuation have been developed. However, existing methods are limited, especially when using real land seismic data. For example, when ground roll and reflections overlap in the time or frequency domains, traditional methods cannot completely separate them and they often distort the signals during the suppression process. We have developed a generative adversarial network (GAN) to attenuate ground roll in seismic data. Unlike traditional methods for ground-roll attenuation dependent on various filters, the GAN method is based on a large training data set that includes pairs of data with and without ground roll. After training the neural network with the training data, the network can identify and filter out any noise in the data. To fulfill this purpose, the proposed method uses a generator and a discriminator. Through network training, the generator learns to create the data that can fool the discriminator, and the discriminator can then distinguish between the data produced by the generator and the training data. As a result of the competition between generators and discriminators, generators produce better images whereas discriminators accurately recognize targets. Tests on synthetic and real land seismic data show that the proposed method effectively reveals reflections masked by the ground roll and obtains better results in the attenuation of ground roll and in the preservation of signals compared to the three other methods.
18
Gazizov, Marat Rushanovich, and Karen Albertovich Grigorian. "Of Neural Network Model Robustness Through Generating Invariant to Attributes Embeddings." Russian Digital Libraries Journal 23, no. 6 (June 4, 2020): 1142–54. http://dx.doi.org/10.26907/1562-5419-2020-23-6-1142-1154.
Повний текст джерелаАнотація:
Model robustness to minor deviations in the distribution of input data is an important criterion in many tasks. Neural networks show high accuracy on training samples, but the quality on test samples can be dropped dramatically due to different data distributions, a situation that is exacerbated at the subgroup level within each category. In this article we show how the robustness of the model at the subgroup level can be significantly improved with the help of the domain adaptation approach to image embeddings. We have found that application of a competitive approach to embeddings limitation gives a significant increase of accuracy metrics in a complex subgroup in comparison with the previous models. The method was tested on two independent datasets, the accuracy in a complex subgroup on the Waterbirds dataset is 90.3 {y : waterbirds;a : landbackground}, on the CelebA dataset is 92.22 {y : blondhair;a : male}.
19
Wang, Dong, Ying Li, Li Ma, Zongwen Bai, and Jonathan Chan. "Going Deeper with Densely Connected Convolutional Neural Networks for Multispectral Pansharpening." Remote Sensing 11, no. 22 (November 7, 2019): 2608. http://dx.doi.org/10.3390/rs11222608.
Повний текст джерелаАнотація:
In recent years, convolutional neural networks (CNNs) have shown promising performance in the field of multispectral (MS) and panchromatic (PAN) image fusion (MS pansharpening). However, the small-scale data and the gradient vanishing problem have been preventing the existing CNN-based fusion approaches from leveraging deeper networks that potentially have better representation ability to characterize the complex nonlinear mapping relationship between the input (source) and the targeting (fused) images. In this paper, we introduce a very deep network with dense blocks and residual learning to tackle these problems. The proposed network takes advantage of dense connections in dense blocks that have connections for arbitrarily two convolution layers to facilitate gradient flow and implicit deep supervision during training. In addition, reusing feature maps can reduce the number of parameters, which is helpful for reducing overfitting that resulted from small-scale data. Residual learning is explored to reduce the difficulty for the model to generate the MS image with high spatial resolution. The proposed network is evaluated via experiments on three datasets, achieving competitive or superior performance, e.g. the spectral angle mapper (SAM) is decreased over 10% on GaoFen-2, when compared with other state-of-the-art methods.
20
Snyder, Hannah R., Marie T. Banich, and Yuko Munakata. "All Competition Is Not Alike: Neural Mechanisms for Resolving Underdetermined and Prepotent Competition." Journal of Cognitive Neuroscience 26, no. 11 (November 2014): 2608–23. http://dx.doi.org/10.1162/jocn_a_00652.
Повний текст джерелаАнотація:
People must constantly select among potential thoughts and actions in the face of competition from (a) multiple task-relevant options (underdetermined competition) and (b) strongly dominant options that are not appropriate in the current context (prepotent competition). These types of competition are ubiquitous during language production. In this work, we investigate the neural mechanisms that allow individuals to effectively manage these cognitive control demands and to quickly choose words with few errors. Using fMRI, we directly contrast underdetermined and prepotent competition within the same task (verb generation) for the first time, allowing localization of the neural substrates supporting the resolution of these two types of competition. Using a neural network model, we investigate the possible mechanisms by which these brain regions support selection. Together, our findings demonstrate that all competition is not alike: resolving prepotent competition and resolving underdetermined competition rely on partly dissociable neural substrates and mechanisms. Specifically, activation of left ventrolateral pFC is specific to resolving underdetermined competition between multiple appropriate responses, most likely via competitive lateral inhibition. In contrast, activation of left dorsolateral pFC is sensitive to both underdetermined competition and prepotent competition from response options that are inappropriate in the current context. This region likely provides top–down support for task-relevant responses, which enables them to out-compete prepotent responses in the selection process that occurs in left ventrolateral pFC.
21
Przednowek, Krzysztof, Janusz Iskra, Krzysztof Wiktorowicz, Tomasz Krzeszowski, and Adam Maszczyk. "Planning Training Loads for The 400 M Hurdles in Three-Month Mesocycles Using Artificial Neural Networks." Journal of Human Kinetics 60, no. 1 (December 28, 2017): 175–89. http://dx.doi.org/10.1515/hukin-2017-0101.
Повний текст джерелаАнотація:
Abstract This paper presents a novel approach to planning training loads in hurdling using artificial neural networks. The neural models performed the task of generating loads for athletes’ training for the 400 meters hurdles. All the models were calculated based on the training data of 21 Polish National Team hurdlers, aged 22.25 ± 1.96, competing between 1989 and 2012. The analysis included 144 training plans that represented different stages in the annual training cycle. The main contribution of this paper is to develop neural models for planning training loads for the entire career of a typical hurdler. In the models, 29 variables were used, where four characterized the runner and 25 described the training process. Two artificial neural networks were used: a multi-layer perceptron and a network with radial basis functions. To assess the quality of the models, the leave-one-out cross-validation method was used in which the Normalized Root Mean Squared Error was calculated. The analysis shows that the method generating the smallest error was the radial basis function network with nine neurons in the hidden layer. Most of the calculated training loads demonstrated a non-linear relationship across the entire competitive period. The resulting model can be used as a tool to assist a coach in planning training loads during a selected training period.
22
Li, Danhua, Xiaofeng Di, Xuan Qu, Yunfei Zhao, and Honggang Kong. "Deep Convolutional Neural Network for Pedestrian Detection with Multi-Levels Features Fusion." MATEC Web of Conferences 232 (2018): 01061. http://dx.doi.org/10.1051/matecconf/201823201061.
Повний текст джерелаАнотація:
Pedestrian detection aims to localize and recognize every pedestrian instance in an image with a bounding box. The current state-of-the-art method is Faster RCNN, which is such a network that uses a region proposal network (RPN) to generate high quality region proposals, while Fast RCNN is used to classifiers extract features into corresponding categories. The contribution of this paper is integrated low-level features and high-level features into a Faster RCNN-based pedestrian detection framework, which efficiently increase the capacity of the feature. Through our experiments, we comprehensively evaluate our framework, on the Caltech pedestrian detection benchmark and our methods achieve state-of-the-art accuracy and present a competitive result on Caltech dataset.
23
Gan, Fengjiao, Chenggao Luo, Xingyue Liu, Hongqiang Wang, and Long Peng. "Fast Terahertz Coded-Aperture Imaging Based on Convolutional Neural Network." Applied Sciences 10, no. 8 (April 12, 2020): 2661. http://dx.doi.org/10.3390/app10082661.
Повний текст джерелаАнотація:
Terahertz coded-aperture imaging (TCAI) has many advantages such as forward-looking imaging, staring imaging and low cost and so forth. However, it is difficult to resolve the target under low signal-to-noise ratio (SNR) and the imaging process is time-consuming. Here, we provide an efficient solution to tackle this problem. A convolution neural network (CNN) is leveraged to develop an off-line end to end imaging network whose structure is highly parallel and free of iterations. And it can just act as a general and powerful mapping function. Once the network is well trained and adopted for TCAI signal processing, the target of interest can be recovered immediately from echo signal. Also, the method to generate training data is shown, and we find that the imaging network trained with simulation data is of good robustness against noise and model errors. The feasibility of the proposed approach is verified by simulation experiments and the results show that it has a competitive performance with the state-of-the-art algorithms.
24
Kanazawa, Yusuke, Tetsuya Asai, and Yoshihito Amemiya. "Basic Circuit Design of a Neural Processor: Analog CMOS Implementation of Spiking Neurons and Dynamic Synapses." Journal of Robotics and Mechatronics 15, no. 2 (April 20, 2003): 208–18. http://dx.doi.org/10.20965/jrm.2003.p0208.
Повний текст джерелаАнотація:
We discuss the integration architecture of spiking neurons, predicted to be next-generation basic circuits of neural processor and dynamic synapse circuits. A key to development of a brain-like processor is to learn from the brain. Learning from the brain, we try to develop circuits implementing neuron and synapse functions while enabling large-scale integration, so large-scale integrated circuits (LSIs) realize functional behavior of neural networks. With such VLSI, we try to construct a large-scale neural network on a single semiconductor chip. With circuit integration now reaching micron levels, however, problems have arisen in dispersion of device performance in analog IC and in the influence of electromagnetic noise. A genuine brain computer should solve such problems on the network level rather than the element level. To achieve such a target, we must develop an architecture that learns brain functions sufficiently and works correctly even in a noisy environment. As the first step, we propose an analog circuit architecture of spiking neurons and dynamic synapses representing the model of artificial neurons and synapses in a form closer to that of the brain. With the proposed circuit, the model of neurons and synapses can be integrated on a silicon chip with metal-oxide-semiconductor (MOS) devices. In the sections that follow, we discuss the dynamic performance of the proposed circuit by using a circuit simulator, HSPICE. As examples of networks using these circuits, we introduce a competitive neural network and an active pattern recognition network by extracting firing frequency information from input information. We also show simulation results of the operation of networks constructed with the proposed circuits.
25
AKHAND, M. A. H., and K. MURASE. "ENSEMBLES OF NEURAL NETWORKS BASED ON THE ALTERATION OF INPUT FEATURE VALUES." International Journal of Neural Systems 22, no. 01 (February 2012): 77–87. http://dx.doi.org/10.1142/s0129065712003079.
Повний текст джерелаАнотація:
An ensemble performs well when the component classifiers are diverse yet accurate, so that the failure of one is compensated for by others. A number of methods have been investigated for constructing ensemble in which some of them train classifiers with the generated patterns. This study investigates a new technique of training pattern generation. The method alters input feature values of some patterns using the values of other patterns to generate different patterns for different classifiers. The effectiveness of neural network ensemble based on the proposed technique was evaluated using a suite of 25 benchmark classification problems, and was found to achieve performance better than or competitive with related conventional methods. Experimental investigation of different input values alteration techniques finds that alteration with pattern values in the same class is better for generalization, although other alteration techniques may offer more diversity.
26
Yang, Zhile, Monjur Mourshed, Kailong Liu, Xinzhi Xu, and Shengzhong Feng. "A novel competitive swarm optimized RBF neural network model for short-term solar power generation forecasting." Neurocomputing 397 (July 2020): 415–21. http://dx.doi.org/10.1016/j.neucom.2019.09.110.
Повний текст джерела
27
Nguyen, Binh, Binh Le, Long H. B. Nguyen, and Dien Dinh. "PhraseAttn: Dynamic Slot Capsule Networks for phrase representation in Neural Machine Translation." Journal of Intelligent & Fuzzy Systems 42, no. 4 (March 4, 2022): 3871–78. http://dx.doi.org/10.3233/jifs-212101.
Повний текст джерелаАнотація:
Word representation plays a vital role in most Natural Language Processing systems, especially for Neural Machine Translation. It tends to capture semantic and similarity between individual words well, but struggle to represent the meaning of phrases or multi-word expressions. In this paper, we investigate a method to generate and use phrase information in a translation model. To generate phrase representations, a Primary Phrase Capsule network is first employed, then iteratively enhancing with a Slot Attention mechanism. Experiments on the IWSLT English to Vietnamese, French, and German datasets show that our proposed method consistently outperforms the baseline Transformer, and attains competitive results over the scaled Transformer with two times lower parameters.
28
Chaushev, Alexander, Liam Raynard, Michael R. Goad, Philipp Eigmüller, David J. Armstrong, Joshua T. Briegal, Matthew R. Burleigh, et al. "Classifying exoplanet candidates with convolutional neural networks: application to the Next Generation Transit Survey." Monthly Notices of the Royal Astronomical Society 488, no. 4 (July 31, 2019): 5232–50. http://dx.doi.org/10.1093/mnras/stz2058.
Повний текст джерелаАнотація:
ABSTRACT Vetting of exoplanet candidates in transit surveys is a manual process, which suffers from a large number of false positives and a lack of consistency. Previous work has shown that convolutional neural networks (CNN) provide an efficient solution to these problems. Here, we apply a CNN to classify planet candidates from the Next Generation Transit Survey (NGTS). For training data sets we compare both real data with injected planetary transits and fully simulated data, as well as how their different compositions affect network performance. We show that fewer hand labelled light curves can be utilized, while still achieving competitive results. With our best model, we achieve an area under the curve (AUC) score of $(95.6\pm {0.2}){{\ \rm per\ cent}}$ and an accuracy of $(88.5\pm {0.3}){{\ \rm per\ cent}}$ on our unseen test data, as well as $(76.5\pm {0.4}){{\ \rm per\ cent}}$ and $(74.6\pm {1.1}){{\ \rm per\ cent}}$ in comparison to our existing manual classifications. The neural network recovers 13 out of 14 confirmed planets observed by NGTS, with high probability. We use simulated data to show that the overall network performance is resilient to mislabelling of the training data set, a problem that might arise due to unidentified, low signal-to-noise transits. Using a CNN, the time required for vetting can be reduced by half, while still recovering the vast majority of manually flagged candidates. In addition, we identify many new candidates with high probabilities which were not flagged by human vetters.
29
Zhang, Xun, Lanyan Yang, Bin Zhang, Ying Liu, Dong Jiang, Xiaohai Qin, and Mengmeng Hao. "Multi-Scale Aggregation Graph Neural Networks Based on Feature Similarity for Semi-Supervised Learning." Entropy 23, no. 4 (March 28, 2021): 403. http://dx.doi.org/10.3390/e23040403.
Повний текст джерелаАнотація:
The problem of extracting meaningful data through graph analysis spans a range of different fields, such as social networks, knowledge graphs, citation networks, the World Wide Web, and so on. As increasingly structured data become available, the importance of being able to effectively mine and learn from such data continues to grow. In this paper, we propose the multi-scale aggregation graph neural network based on feature similarity (MAGN), a novel graph neural network defined in the vertex domain. Our model provides a simple and general semi-supervised learning method for graph-structured data, in which only a very small part of the data is labeled as the training set. We first construct a similarity matrix by calculating the similarity of original features between all adjacent node pairs, and then generate a set of feature extractors utilizing the similarity matrix to perform multi-scale feature propagation on graphs. The output of multi-scale feature propagation is finally aggregated by using the mean-pooling operation. Our method aims to improve the model representation ability via multi-scale neighborhood aggregation based on feature similarity. Extensive experimental evaluation on various open benchmarks shows the competitive performance of our method compared to a variety of popular architectures.
30
Zhang, Kai, Guanghua Xu, Zezhen Han, Kaiquan Ma, Xiaowei Zheng, Longting Chen, Nan Duan, and Sicong Zhang. "Data Augmentation for Motor Imagery Signal Classification Based on a Hybrid Neural Network." Sensors 20, no. 16 (August 11, 2020): 4485. http://dx.doi.org/10.3390/s20164485.
Повний текст джерелаАнотація:
As an important paradigm of spontaneous brain-computer interfaces (BCIs), motor imagery (MI) has been widely used in the fields of neurological rehabilitation and robot control. Recently, researchers have proposed various methods for feature extraction and classification based on MI signals. The decoding model based on deep neural networks (DNNs) has attracted significant attention in the field of MI signal processing. Due to the strict requirements for subjects and experimental environments, it is difficult to collect large-scale and high-quality electroencephalogram (EEG) data. However, the performance of a deep learning model depends directly on the size of the datasets. Therefore, the decoding of MI-EEG signals based on a DNN has proven highly challenging in practice. Based on this, we investigated the performance of different data augmentation (DA) methods for the classification of MI data using a DNN. First, we transformed the time series signals into spectrogram images using a short-time Fourier transform (STFT). Then, we evaluated and compared the performance of different DA methods for this spectrogram data. Next, we developed a convolutional neural network (CNN) to classify the MI signals and compared the classification performance of after DA. The Fréchet inception distance (FID) was used to evaluate the quality of the generated data (GD) and the classification accuracy, and mean kappa values were used to explore the best CNN-DA method. In addition, analysis of variance (ANOVA) and paired t-tests were used to assess the significance of the results. The results showed that the deep convolutional generative adversarial network (DCGAN) provided better augmentation performance than traditional DA methods: geometric transformation (GT), autoencoder (AE), and variational autoencoder (VAE) (p < 0.01). Public datasets of the BCI competition IV (datasets 1 and 2b) were used to verify the classification performance. Improvements in the classification accuracies of 17% and 21% (p < 0.01) were observed after DA for the two datasets. In addition, the hybrid network CNN-DCGAN outperformed the other classification methods, with average kappa values of 0.564 and 0.677 for the two datasets.
31
Lee, Byung-Jun, and Kee-Eung Kim. "Dialog History Construction with Long-Short Term Memory for Robust Generative Dialog State Tracking." Dialogue & Discourse 7, no. 3 (April 15, 2016): 47–64. http://dx.doi.org/10.5087/dad.2016.302.
Повний текст джерелаАнотація:
One of the crucial components of dialog system is the dialog state tracker, which infers user’s intention from preliminary speech processing. Since the overall performance of the dialog system is heavily affected by that of the dialog tracker, it has been one of the core areas of research on dialog systems. In this paper, we present a dialog state tracker that combines a generative probabilistic model of dialog state tracking with the recurrent neural network for encoding important aspects of the dialog history. We describe a two-step gradient descent algorithm that optimizes the tracker with a complex loss function. We demonstrate that this approach yields a dialog state tracker that performs competitively with top-performing trackers participated in the first and second Dialog State Tracking Challenges.
32
Peng, Xiaoming, Abdesselam Bouzerdoum, and Son Lam Phung. "A Trajectory-Based Method for Dynamic Scene Recognition." International Journal of Pattern Recognition and Artificial Intelligence 35, no. 10 (May 15, 2021): 2150029. http://dx.doi.org/10.1142/s0218001421500294.
Повний текст джерелаАнотація:
Existing methods for dynamic scene recognition mostly use global features extracted from the entire video frame or a video segment. In this paper, a trajectory-based dynamic scene recognition method is proposed. A trajectory is formed by a pixel moving across consecutive frames of a video segment. The local regions surrounding the trajectory provide useful appearance and motion information about a portion of the video segment. The proposed method works at several stages. First, dense and evenly distributed trajectories are extracted from a video segment. Then, the fully-connected-layer features are extracted from each trajectory using a pre-trained Convolutional Neural Networks (CNNs) model, forming a feature sequence. Next, these feature sequences are fed into a Long-Short-Term-Memory (LSTM) network to learn their temporal behavior. Finally, by aggregating the information of the trajectories, a global representation of the video segment can be obtained for classification purposes. The LSTM is trained using synthetic trajectory feature sequences instead of real ones. The synthetic feature sequences are generated with a series of generative adversarial networks (GANs). In addition to classification, category-specific discriminative trajectories are located in a video segment, which help reveal what portions of a video segment are more important than others. This is achieved by formulating an optimization problem to learn discriminative part detectors for all categories simultaneously. Experimental results on two benchmark dynamic scene datasets show that the proposed method is very competitive with six other methods.
33
Mai, Weimin, Junxin Chen, and Xiang Chen. "Time-Evolving Graph Convolutional Recurrent Network for Traffic Prediction." Applied Sciences 12, no. 6 (March 10, 2022): 2842. http://dx.doi.org/10.3390/app12062842.
Повний текст джерелаАнотація:
Accurate traffic prediction is crucial to the construction of intelligent transportation systems. This task remains challenging because of the complicated and dynamic spatiotemporal dependency in traffic networks. While various graph-based spatiotemporal networks have been proposed for traffic prediction, most of them rely on predefined graphs from different views or static adaptive matrices. Some implicit dynamics of inter-node dependency may be neglected, which limits the performance of prediction. To address this problem and make more accurate predictions, we propose a traffic prediction model named Time-Evolving Graph Convolution Recurrent Network (TEGCRN), which takes advantage of time-evolving graph convolution to capture the dynamic inter-node dependency adaptively at different time slots. Specifically, we first propose a tensor-composing method to generate adaptive time-evolving adjacency graphs. Based on these time-evolving graphs and a predefined distance-based graph, a graph convolution module with mix-hop operation is applied to extract comprehensive inter-node information. Then the resulting graph convolution module is integrated into the Recurrent Neural Network structure to form an general predicting model. Experiments on two real-world traffic datasets demonstrate the superiority of TEGCRN over multiple competitive baseline models, especially in short-term prediction, which also verifies the effectiveness of time-evolving graph convolution in capturing more comprehensive inter-node dependency.
34
Lv, Gang, Yushan Xu, Zuchang Ma, Yining Sun, and Fudong Nian. "Scale and Background Aware Asymmetric Bilateral Network for Unconstrained Image Crowd Counting." Mathematics 10, no. 7 (March 25, 2022): 1053. http://dx.doi.org/10.3390/math10071053.
Повний текст джерелаАнотація:
This paper attacks the two challenging problems of image-based crowd counting, that is, scale variation and complex background. To that end, we present a novel crowd counting method, called the Scale and Background aware Asymmetric Bilateral Network (SBAB-Net), which is able to handle scale variation and background noise in a unified framework. Specifically, the proposed SBAB-Net contains three main components, a pre-trained backbone convolutional neural network (CNN) as the feature extractor and two asymmetric branches to generate a density map. These two asymmetric branches have different structures and use features from different semantic layers. One branch is densely connected stacked dilated convolution (DCSDC) sub-network with different dilation rates, which relies on one deep feature layer and can handle scale variation. The other branch is parameter-free densely connected stacked pooling (DCSP) sub-network with various pooling kernels and strides, which relies on shallow feature and can fuse features with several receptive fields to reduce the impact of background noise. Two sub-networks are fused by the attention mechanism to generate the final density map. Extensive experimental results on three widely-used benchmark datasets have demonstrated the effectiveness and superiority of our proposed method: (1) We achieve competitive counting performance compared to state-of-the-art methods; (2) Compared with baseline, the MAE and MSE are decreased by at least 6.3% and 11.3%, respectively.
35
Bologna, Guido, and Yoichi Hayashi. "A Comparison Study on Rule Extraction from Neural Network Ensembles, Boosted Shallow Trees, and SVMs." Applied Computational Intelligence and Soft Computing 2018 (2018): 1–20. http://dx.doi.org/10.1155/2018/4084850.
Повний текст джерелаАнотація:
One way to make the knowledge stored in an artificial neural network more intelligible is to extract symbolic rules. However, producing rules from Multilayer Perceptrons (MLPs) is an NP-hard problem. Many techniques have been introduced to generate rules from single neural networks, but very few were proposed for ensembles. Moreover, experiments were rarely assessed by 10-fold cross-validation trials. In this work, based on the Discretized Interpretable Multilayer Perceptron (DIMLP), experiments were performed on 10 repetitions of stratified 10-fold cross-validation trials over 25 binary classification problems. The DIMLP architecture allowed us to produce rules from DIMLP ensembles, boosted shallow trees (BSTs), and Support Vector Machines (SVM). The complexity of rulesets was measured with the average number of generated rules and average number of antecedents per rule. From the 25 used classification problems, the most complex rulesets were generated from BSTs trained by “gentle boosting” and “real boosting.” Moreover, we clearly observed that the less complex the rules were, the better their fidelity was. In fact, rules generated from decision stumps trained by modest boosting were, for almost all the 25 datasets, the simplest with the highest fidelity. Finally, in terms of average predictive accuracy and average ruleset complexity, the comparison of some of our results to those reported in the literature proved to be competitive.
36
Dai, Wenxin, Yuqing Mao, Rongao Yuan, Yijing Liu, Xuemei Pu, and Chuan Li. "A Novel Detector Based on Convolution Neural Networks for Multiscale SAR Ship Detection in Complex Background." Sensors 20, no. 9 (April 30, 2020): 2547. http://dx.doi.org/10.3390/s20092547.
Повний текст джерелаАнотація:
Convolution neural network (CNN)-based detectors have shown great performance on ship detections of synthetic aperture radar (SAR) images. However, the performance of current models has not been satisfactory enough for detecting multiscale ships and small-size ones in front of complex backgrounds. To address the problem, we propose a novel SAR ship detector based on CNN, which consist of three subnetworks: the Fusion Feature Extractor Network (FFEN), Region Proposal Network (RPN), and Refine Detection Network (RDN). Instead of using a single feature map, we fuse feature maps in bottom–up and top–down ways and generate proposals from each fused feature map in FFEN. Furthermore, we further merge features generated by the region-of-interest (RoI) pooling layer in RDN. Based on the feature representation strategy, the CNN framework constructed can significantly enhance the location and semantics information for the multiscale ships, in particular for the small ships. On the other hand, the residual block is introduced to increase the network depth, through which the detection precision could be further improved. The public SAR ship dataset (SSDD) and China Gaofen-3 satellite SAR image are used to validate the proposed method. Our method shows excellent performance for detecting the multiscale and small-size ships with respect to some competitive models and exhibits high potential in practical application.
37
Piao, Jingchun, Yunfan Chen, and Hyunchul Shin. "A New Deep Learning Based Multi-Spectral Image Fusion Method." Entropy 21, no. 6 (June 5, 2019): 570. http://dx.doi.org/10.3390/e21060570.
Повний текст джерелаАнотація:
In this paper, we present a new effective infrared (IR) and visible (VIS) image fusion method by using a deep neural network. In our method, a Siamese convolutional neural network (CNN) is applied to automatically generate a weight map which represents the saliency of each pixel for a pair of source images. A CNN plays a role in automatic encoding an image into a feature domain for classification. By applying the proposed method, the key problems in image fusion, which are the activity level measurement and fusion rule design, can be figured out in one shot. The fusion is carried out through the multi-scale image decomposition based on wavelet transform, and the reconstruction result is more perceptual to a human visual system. In addition, the visual qualitative effectiveness of the proposed fusion method is evaluated by comparing pedestrian detection results with other methods, by using the YOLOv3 object detector using a public benchmark dataset. The experimental results show that our proposed method showed competitive results in terms of both quantitative assessment and visual quality.
38
Ma, Xianghua, Zhenkun Yang, and Shining Chen. "Multiscale Feature Filtering Network for Image Recognition System in Unmanned Aerial Vehicle." Complexity 2021 (February 18, 2021): 1–11. http://dx.doi.org/10.1155/2021/6663851.
Повний текст джерелаАнотація:
For unmanned aerial vehicle (UAV), object detection at different scales is an important component for the visual recognition. Recent advances in convolutional neural networks (CNNs) have demonstrated that attention mechanism remarkably enhances multiscale representation of CNNs. However, most existing multiscale feature representation methods simply employ several attention blocks in the attention mechanism to adaptively recalibrate the feature response, which overlooks the context information at a multiscale level. To solve this problem, a multiscale feature filtering network (MFFNet) is proposed in this paper for image recognition system in the UAV. A novel building block, namely, multiscale feature filtering (MFF) module, is proposed for ResNet-like backbones and it allows feature-selective learning for multiscale context information across multiparallel branches. These branches employ multiple atrous convolutions at different scales, respectively, and further adaptively generate channel-wise feature responses by emphasizing channel-wise dependencies. Experimental results on CIFAR100 and Tiny ImageNet datasets reflect that the MFFNet achieves very competitive results in comparison with previous baseline models. Further ablation experiments verify that the MFFNet can achieve consistent performance gains in image classification and object detection tasks.
39
Tan, Kok Sheng, and Preethi Subramanian. "Proposition of Machine Learning Driven Personalized Marketing Approach for E-Commerce." Journal of Computational and Theoretical Nanoscience 16, no. 8 (August 1, 2019): 3532–37. http://dx.doi.org/10.1166/jctn.2019.8319.
Повний текст джерелаАнотація:
The ubiquity of digital devices and Internet has formed a constantly connected online environment which led to the extensive adoption of e-commerce. However, the active participation of growing number of stakeholders intensifies the highly competitive landscape of the dynamic e-commerce market and the scarcity of trust in e-commerce business impede the generation of consistent sales growth. The obstruction necessitates the implementation of innovative marketing strategies to enhance the relationships with customers to develop customer loyalty. Therefore, a machine learning driven personalized marketing approach is proposed to facilitate the implementation of personalized marketing in which there are 2 significant sequential elements namely, the development of personalized marketing contents and delivery of the contents to prospective customers. Cluster analysis is employed to perform customer segmentation to discover customer segments due to the capability of the analysis to identify similarities in customer preferences in which the discovered customer segments are used to construct personalized marketing contents. In addition, artificial neural network is employed to predict prospective customers due to the capability of artificial neural network to comprehend complex relationships between customer demographics and buying behaviour in which the prediction facilitates the delivery of the constructed personalized marketing contents to potential repeat customer to optimize the marketing initiative. The combination of cluster analysis and artificial neural network empowers the construction of an efficacious marketing pipeline which enhances the competency of e-commerce businesses.
40
He, Minxue, Liheng Zhong, Prabhjot Sandhu, and Yu Zhou. "Emulation of a Process-Based Salinity Generator for the Sacramento–San Joaquin Delta of California via Deep Learning." Water 12, no. 8 (July 23, 2020): 2088. http://dx.doi.org/10.3390/w12082088.
Повний текст джерелаАнотація:
Salinity management is a subject of particular interest in estuarine environments because of the underlying biological significance of salinity and its variations in time and space. The foremost step in such management practices is understanding the spatial and temporal variations of salinity and the principal drivers of these variations. This has traditionally been achieved with the assistance of empirical or process-based models, but these can be computationally expensive for complex environmental systems. Model emulation based on data-driven methods offers a viable alternative to traditional modeling in terms of computational efficiency and improving accuracy by recognizing patterns and processes that are overlooked or underrepresented (or overrepresented) by traditional models. This paper presents a case study of emulating a process-based boundary salinity generator via deep learning for the Sacramento–San Joaquin Delta (Delta), an estuarine environment with significant economic, ecological, and social value on the Pacific coast of northern California, United States. Specifically, the study proposes a range of neural network models: (a) multilayer perceptron, (b) long short-term memory network, and (c) convolutional neural network-based models in estimating the downstream boundary salinity of the Delta on a daily time-step. These neural network models are trained and validated using half of the dataset from water year 1991 to 2002. They are then evaluated for performance in the remaining record period from water year 2003 to 2014 against the process-based boundary salinity generation model across different ranges of salinity in different types of water years. The results indicate that deep learning neural networks provide competitive or superior results compared with the process-based model, particularly when the output of the latter are incorporated as an input to the former. The improvements are generally more noticeable during extreme (i.e., wet, dry, and critical) years rather than in near-normal (i.e., above-normal and below-normal) years and during low and medium ranges of salinity rather than high range salinity. Overall, this study indicates that deep learning approaches have the potential to supplement the current practices in estimating salinity at the downstream boundary and other locations across the Delta, and thus guide real-time operations and long-term planning activities in the Delta.
41
Chen, Zuyao, Qianqian Xu, Runmin Cong, and Qingming Huang. "Global Context-Aware Progressive Aggregation Network for Salient Object Detection." Proceedings of the AAAI Conference on Artificial Intelligence 34, no. 07 (April 3, 2020): 10599–606. http://dx.doi.org/10.1609/aaai.v34i07.6633.
Повний текст джерелаАнотація:
Deep convolutional neural networks have achieved competitive performance in salient object detection, in which how to learn effective and comprehensive features plays a critical role. Most of the previous works mainly adopted multiple-level feature integration yet ignored the gap between different features. Besides, there also exists a dilution process of high-level features as they passed on the top-down pathway. To remedy these issues, we propose a novel network named GCPANet to effectively integrate low-level appearance features, high-level semantic features, and global context features through some progressive context-aware Feature Interweaved Aggregation (FIA) modules and generate the saliency map in a supervised way. Moreover, a Head Attention (HA) module is used to reduce information redundancy and enhance the top layers features by leveraging the spatial and channel-wise attention, and the Self Refinement (SR) module is utilized to further refine and heighten the input features. Furthermore, we design the Global Context Flow (GCF) module to generate the global context information at different stages, which aims to learn the relationship among different salient regions and alleviate the dilution effect of high-level features. Experimental results on six benchmark datasets demonstrate that the proposed approach outperforms the state-of-the-art methods both quantitatively and qualitatively.
42
Shu, Yuanjun, Wei Li, Menglong Yang, Peng Cheng, and Songchen Han. "Patch-Based Change Detection Method for SAR Images with Label Updating Strategy." Remote Sensing 13, no. 7 (March 24, 2021): 1236. http://dx.doi.org/10.3390/rs13071236.
Повний текст джерелаАнотація:
Convolutional neural networks (CNNs) have been widely used in change detection of synthetic aperture radar (SAR) images and have been proven to have better precision than traditional methods. A two-stage patch-based deep learning method with a label updating strategy is proposed in this paper. The initial label and mask are generated at the pre-classification stage. Then a two-stage updating strategy is applied to gradually recover changed areas. At the first stage, diversity of training data is gradually restored. The output of the designed CNN network is further processed to generate a new label and a new mask for the following learning iteration. As the diversity of data is ensured after the first stage, pixels within uncertain areas can be easily classified at the second stage. Experiment results on several representative datasets show the effectiveness of our proposed method compared with several existing competitive methods.
43
Yan, Zheren, Can Yang, Lei Hu, Jing Zhao, Liangcun Jiang, and Jianya Gong. "The Integration of Linguistic and Geospatial Features Using Global Context Embedding for Automated Text Geocoding." ISPRS International Journal of Geo-Information 10, no. 9 (August 24, 2021): 572. http://dx.doi.org/10.3390/ijgi10090572.
Повний текст джерелаАнотація:
Geocoding is an essential procedure in geographical information retrieval to associate place names with coordinates. Due to the inherent ambiguity of place names in natural language and the scarcity of place names in textual data, it is widely recognized that geocoding is challenging. Recent advances in deep learning have promoted the use of the neural network to improve the performance of geocoding. However, most of the existing approaches consider only the local context, e.g., neighboring words in a sentence, as opposed to the global context, e.g., the topic of the document. Lack of global information may have a severe impact on the robustness of the model. To fill the research gap, this paper proposes a novel global context embedding approach to generate linguistic and geospatial features through topic embedding and location embedding, respectively. A deep neural network called LGGeoCoder, which integrates local and global features, is developed to solve the geocoding as a classification problem. The experiments on a Wikipedia place name dataset demonstrate that LGGeoCoder achieves competitive performance compared with state-of-the-art models. Furthermore, the effect of introducing global linguistic and geospatial features in geocoding to alleviate the ambiguity and scarcity problem is discussed.
44
Yan, Huijiong, Tao Qian, Liang Xie, and Shanguang Chen. "Unsupervised cross-lingual model transfer for named entity recognition with contextualized word representations." PLOS ONE 16, no. 9 (September 21, 2021): e0257230. http://dx.doi.org/10.1371/journal.pone.0257230.
Повний текст джерелаАнотація:
Named entity recognition (NER) is one fundamental task in the natural language processing (NLP) community. Supervised neural network models based on contextualized word representations can achieve highly-competitive performance, which requires a large-scale manually-annotated corpus for training. While for the resource-scarce languages, the construction of such as corpus is always expensive and time-consuming. Thus, unsupervised cross-lingual transfer is one good solution to address the problem. In this work, we investigate the unsupervised cross-lingual NER with model transfer based on contextualized word representations, which greatly advances the cross-lingual NER performance. We study several model transfer settings of the unsupervised cross-lingual NER, including (1) different types of the pretrained transformer-based language models as input, (2) the exploration strategies of the multilingual contextualized word representations, and (3) multi-source adaption. In particular, we propose an adapter-based word representation method combining with parameter generation network (PGN) better to capture the relationship between the source and target languages. We conduct experiments on a benchmark ConLL dataset involving four languages to simulate the cross-lingual setting. Results show that we can obtain highly-competitive performance by cross-lingual model transfer. In particular, our proposed adapter-based PGN model can lead to significant improvements for cross-lingual NER.
45
Marnerides, Demetris, Thomas Bashford-Rogers, and Kurt Debattista. "Deep HDR Hallucination for Inverse Tone Mapping." Sensors 21, no. 12 (June 11, 2021): 4032. http://dx.doi.org/10.3390/s21124032.
Повний текст джерелаАнотація:
Inverse Tone Mapping (ITM) methods attempt to reconstruct High Dynamic Range (HDR) information from Low Dynamic Range (LDR) image content. The dynamic range of well-exposed areas must be expanded and any missing information due to over/under-exposure must be recovered (hallucinated). The majority of methods focus on the former and are relatively successful, while most attempts on the latter are not of sufficient quality, even ones based on Convolutional Neural Networks (CNNs). A major factor for the reduced inpainting quality in some works is the choice of loss function. Work based on Generative Adversarial Networks (GANs) shows promising results for image synthesis and LDR inpainting, suggesting that GAN losses can improve inverse tone mapping results. This work presents a GAN-based method that hallucinates missing information from badly exposed areas in LDR images and compares its efficacy with alternative variations. The proposed method is quantitatively competitive with state-of-the-art inverse tone mapping methods, providing good dynamic range expansion for well-exposed areas and plausible hallucinations for saturated and under-exposed areas. A density-based normalisation method, targeted for HDR content, is also proposed, as well as an HDR data augmentation method targeted for HDR hallucination.
46
Cui, Jianming, Wenxiu Kong, Xiaojun Zhang, Da Chen, and Qingtian Zeng. "DLSTM-Based Successive Cancellation Flipping Decoder for Short Polar Codes." Entropy 23, no. 7 (July 6, 2021): 863. http://dx.doi.org/10.3390/e23070863.
Повний текст джерелаАнотація:
Polar code has been adopted as the control channel coding scheme for the fifth generation (5G), and the performance of short polar codes is receiving intensive attention. The successive cancellation flipping (SC flipping) algorithm suffers a significant performance loss in short block lengths. To address this issue, we propose a double long short-term memory (DLSTM) neural network to locate the first error bit. To enhance the prediction accuracy of the DLSTM network, all frozen bits are clipped in the output layer. Then, Gaussian approximation is applied to measure the channel reliability and rank the flipping set to choose the least reliable position for multi-bit flipping. To be robust under different codewords, padding and masking strategies aid the network architecture to be compatible with multiple block lengths. Numerical results indicate that the error-correction performance of the proposed algorithm is competitive with that of the CA-SCL algorithm. It has better performance than the machine learning-based multi-bit flipping SC (ML-MSCF) decoder and the dynamic SC flipping (DSCF) decoder for short polar codes.
47
Wang, Xuchu, Fusheng Wang, and Yanmin Niu. "A Convolutional Neural Network Combining Discriminative Dictionary Learning and Sequence Tracking for Left Ventricular Detection." Sensors 21, no. 11 (May 26, 2021): 3693. http://dx.doi.org/10.3390/s21113693.
Повний текст джерелаАнотація:
Cardiac MRI left ventricular (LV) detection is frequently employed to assist cardiac registration or segmentation in computer-aided diagnosis of heart diseases. Focusing on the challenging problems in LV detection, such as the large span and varying size of LV areas in MRI, as well as the heterogeneous myocardial and blood pool parts in LV areas, a convolutional neural network (CNN) detection method combining discriminative dictionary learning and sequence tracking is proposed in this paper. To efficiently represent the different sub-objects in LV area, the method deploys discriminant dictionary to classify the superpixel oversegmented regions, then the target LV region is constructed by label merging and multi-scale adaptive anchors are generated in the target region for handling the varying sizes. Combining with non-differential anchors in regional proposal network, the left ventricle object is localized by the CNN based regression and classification strategy. In order to solve the problem of slow classification speed of discriminative dictionary, a fast generation module of left ventricular scale adaptive anchors based on sequence tracking is also proposed on the same individual. The method and its variants were tested on the heart atlas data set. Experimental results verified the effectiveness of the proposed method and according to some evaluation indicators, it obtained 92.95% in AP50 metric and it was the most competitive result compared to typical related methods. The combination of discriminative dictionary learning and scale adaptive anchor improves adaptability of the proposed algorithm to the varying left ventricular areas. This study would be beneficial in some cardiac image processing such as region-of-interest cropping and left ventricle volume measurement.
48
Killestein, T. L., J. Lyman, D. Steeghs, K. Ackley, M. J. Dyer, K. Ulaczyk, R. Cutter, et al. "Transient-optimized real-bogus classification with Bayesian convolutional neural networks – sifting the GOTO candidate stream." Monthly Notices of the Royal Astronomical Society 503, no. 4 (March 15, 2021): 4838–54. http://dx.doi.org/10.1093/mnras/stab633.
Повний текст джерелаАнотація:
ABSTRACT Large-scale sky surveys have played a transformative role in our understanding of astrophysical transients, only made possible by increasingly powerful machine learning-based filtering to accurately sift through the vast quantities of incoming data generated. In this paper, we present a new real-bogus classifier based on a Bayesian convolutional neural network that provides nuanced, uncertainty-aware classification of transient candidates in difference imaging, and demonstrate its application to the datastream from the GOTO wide-field optical survey. Not only are candidates assigned a well-calibrated probability of being real, but also an associated confidence that can be used to prioritize human vetting efforts and inform future model optimization via active learning. To fully realize the potential of this architecture, we present a fully automated training set generation method which requires no human labelling, incorporating a novel data-driven augmentation method to significantly improve the recovery of faint and nuclear transient sources. We achieve competitive classification accuracy (FPR and FNR both below 1 per cent) compared against classifiers trained with fully human-labelled data sets, while being significantly quicker and less labour-intensive to build. This data-driven approach is uniquely scalable to the upcoming challenges and data needs of next-generation transient surveys. We make our data generation and model training codes available to the community.
49
Zhang, Junjie, Zhouyin Cai, Fansheng Chen, and Dan Zeng. "Hyperspectral Image Denoising via Adversarial Learning." Remote Sensing 14, no. 8 (April 7, 2022): 1790. http://dx.doi.org/10.3390/rs14081790.
Повний текст джерелаАнотація:
Due to sensor instability and atmospheric interference, hyperspectral images (HSIs) often suffer from different kinds of noise which degrade the performance of downstream tasks. Therefore, HSI denoising has become an essential part of HSI preprocessing. Traditional methods tend to tackle one specific type of noise and remove it iteratively, resulting in drawbacks including inefficiency when dealing with mixed noise. Most recently, deep neural network-based models, especially generative adversarial networks, have demonstrated promising performance in generic image denoising. However, in contrast to generic RGB images, HSIs often possess abundant spectral information; thus, it is non-trivial to design a denoising network to effectively explore both spatial and spectral characteristics simultaneously. To address the above issues, in this paper, we propose an end-to-end HSI denoising model via adversarial learning. More specifically, to capture the subtle noise distribution from both spatial and spectral dimensions, we designed a Residual Spatial-Spectral Module (RSSM) and embed it in an UNet-like structure as the generator to obtain clean images. To distinguish the real image from the generated one, we designed a discriminator based on the Multiscale Feature Fusion Module (MFFM) to further improve the quality of the denoising results. The generator was trained with joint loss functions, including reconstruction loss, structural loss and adversarial loss. Moreover, considering the lack of publicly available training data for the HSI denoising task, we collected an additional benchmark dataset denoted as the Shandong Feicheng Denoising (SFD) dataset. We evaluated five types of mixed noise across several datasets in comparative experiments, and comprehensive experimental results on both simulated and real data demonstrate that the proposed model achieves competitive results against state-of-the-art methods. For ablation studies, we investigated the structure of the generator as well as the training process with joint losses and different amounts of training data, further validating the rationality and effectiveness of the proposed method.
50
Zafar, Haroon, Junaid Zafar, and Faisal Sharif. "Automated Clinical Decision Support for Coronary Plaques Characterization from Optical Coherence Tomography Imaging with Fused Neural Networks." Optics 3, no. 1 (January 10, 2022): 8–18. http://dx.doi.org/10.3390/opt3010002.
Повний текст джерелаАнотація:
Deep Neural Networks (DNNs) are nurturing clinical decision support systems for the detection and accurate modeling of coronary arterial plaques. However, efficient plaque characterization in time-constrained settings is still an open problem. The purpose of this study is to develop a novel automated classification architecture viable for the real-time clinical detection and classification of coronary artery plaques, and secondly, to use the novel dataset of OCT images for data augmentation. Further, the purpose is to validate the efficacy of transfer learning for arterial plaques classification. In this perspective, a novel time-efficient classification architecture based on DNNs is proposed. A new data set consisting of in-vivo patient Optical Coherence Tomography (OCT) images labeled by three trained experts was created and dynamically programmed. Generative Adversarial Networks (GANs) were used for populating the coronary aerial plaques dataset. We removed the fully connected layers, including softmax and the cross-entropy in the GoogleNet framework, and replaced them with the Support Vector Machines (SVMs). Our proposed architecture limits weight up-gradation cycles to only modified layers and computes the global hyper-plane in a timely, competitive fashion. Transfer learning was used for high-level discriminative feature learning. Cross-entropy loss was minimized by using the Adam optimizer for model training. A train validation scheme was used to determine the classification accuracy. Automated plaques differentiation in addition to their detection was found to agree with the clinical findings. Our customized fused classification scheme outperforms the other leading reported works with an overall accuracy of 96.84%, and multiple folds reduced elapsed time demonstrating it as a viable choice for real-time clinical settings.