• Journal of Korea Multimedia Society
• /
• v.25 no.1
• /
• pp.52-60
• /
• 2022

Few-shot learning is a sub-area of machine learning problems, which aims to classify target images that only contain a few labeled samples for training. As a representative few-shot learning method, the Prototypical network has been received much attention due to its simplicity and promising results. However, the Prototypical network uses the sample mean of samples from the same class as the prototypes of that class, which easily results in learning uncharacteristic features in the low-data scenery. In this study, we propose to use local descriptors (i.e., patches along the channel within feature maps) from the same class to explicitly obtain more representative prototypes for Prototypical Network so that significant intra-class feature information can be maintained and thus improving the classification performance on few-shot learning tasks. Experimental results on various benchmark datasets including mini-ImageNet, CUB-200-2011, and tiered-ImageNet show that the proposed method can learn more discriminative intra-class features by the local descriptors and obtain more generic prototype representations under the few-shot setting.

• The Korean Journal of Food And Nutrition
• /
• v.9 no.2
• /
• pp.213-216
• /
• 1996

A modified method is given for the precolumn derivatization and subsequent high-pressure liquid chromatographic seperation of 3-methylhistidine from urine. The elution contained isocratic solution with acetonirile and 10 M sodium phosphate(pH 7.5) requires less than 7 min. The recoveries of 3-methylhlstidine from urine control were 95% to 106%. 3-Methylhistidine determinations were performed on urine samples from volunteers who were both male trained and non-trained physical undergraduates. As the result, urinary .3-methylhistidine content of volunteers increased significantly after weight training.

• Communications for Statistical Applications and Methods
• /
• v.16 no.4
• /
• pp.675-686
• /
• 2009

In this paper, it will be assumed that there are two distinct populations which are multivariate normal with equal covariance matrix. We also assume that the two populations are equally likely and the costs of misclassification are equal. The classification rule depends on the situation when the training samples include missing values or not. We consider the bootstrap confidence intervals for classification error rate when a block of observation is missing.

• Communications for Statistical Applications and Methods
• /
• v.16 no.5
• /
• pp.881-889
• /
• 2009

Support vector machines(SVMs) are known as flexible and efficient classifier of multivariate observations, producing a hyperplane or hyperdimensional curved surface in multidimensional feature space that best separates training samples by known groups. As various methodological extensions are made for SVM classifiers in recent years, it becomes more difficult to understand the constructed model intuitively. The aim of this paper is to visualize various SVM classifications tuned by several parameters in reduced dimensions, so that data analysts secure the tangible image of the products that the machine made.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1995.06a
• /
• pp.184-187
• /
• 1995

This paper proposes an algorithm for machine recognition of phonemes in continuous speech. The proposed algorithm is static strategy neural network. The algorithm uses, at the stage of training neuron, features such as PARCOR coefficient and auditory-like perceptual liner prediction . These features are extracted from speech samples selected by a sliding 25.6msec windows with s sliding gap being 3 msec long, then interleaved and summed up to 7 sets of parmeters covering 171 msec worth of speech for use of neural inputs. Perfomances are compared when either PARCOR or auditory-like PLP is included in the feture set.

• Speech Sciences
• /
• v.12 no.4
• /
• pp.215-226
• /
• 2005

This study is an attempt to investigate effect of pitch treatment program using visual feedback for profound deaf adults. Dr. Speech program was applied as a training tool. The subjects of this study were 3 profound deaf adults. Speech samples for evaluation were vowel prolongations and connected speech. Analysis was performed under the principle of single subject research design. As results of this study, all subjects showed the treatment effects which were represented by lowering fundamental frequency and speaking fundamental frequency.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.10 no.12
• /
• pp.5643-5656
• /
• 2016

The detection accuracy of steganalysis depends on many factors, including the embedding algorithm, the payload size, the steganalysis feature space and the properties of the cover source. In practice, the cover source mismatch (CSM) problem has been recognized as the single most important factor negatively affecting the performance. To address this problem, we propose a new framework for blind, universal steganalysis which uses traditional steganalyst features. Firstly, cover images with the same statistical properties are searched from a reference image database as aided samples. The test image and its aided samples form a whole test set. Then, by assuming that most of the aided samples are innocent, we conduct outlier detection on the test set to judge the test image as cover or stego. In this way, the framework has removed the need for training. Hence, it does not suffer from cover source mismatch. Because it performs anomaly detection rather than classification, this method is totally unsupervised. The results in our study show that this framework works superior than one-class support vector machine and the outlier detector without considering the image retrieval process.

• Computers and Concrete
• /
• v.15 no.2
• /
• pp.167-181
• /
• 2015

"Mortar or concrete pneumatically projected at high velocity onto a surface" is called Shotcrete. Models that predict shotcrete design parameters (e.g. compressive strength, slump etc) from any mixing proportions of admixtures could save considerable experimentation time consumed during trial and error based procedures. Artificial Neural Network (ANN) has been widely used for similar purposes; however, such models have been rarely applied on shotcrete design. In this study 19 samples of shotcrete test panels with varying quantities of water, steel fibers and silica fume were used to determine their slump, cost and compressive strength at different ages. A number of 3-layer Back propagation Neural Network (BPNN) models of different network architectures were used to train the network using 15 samples, while 4 samples were randomly chosen to validate the model. The predicted compressive strength from linear regression lacked accuracy with $R^2$ value of 0.36. Whereas, outputs from 3-5-3 ANN architecture gave higher correlations of $R^2$ = 0.99, 0.95 and 0.98 for compressive strength, cost and slump parameters of the training data and corresponding $R^2$ values of 0.99, 0.99 and 0.90 for the validation dataset. Sensitivity analysis of output variables using ANN can unfold the nonlinear cause and effect relationship for otherwise obscure ANN model.

• Water for future
• /
• v.13 no.3
• /
• pp.71-76
• /
• 1980

The oxidation ditch is an efficient, low cost form of treatment of domestic and many industrial waste. It has gained rapid acceptance because of its simplicity, low cost operation, ease of operation, simple mintenance and flexibiltiy. The objective of this investigation was to measure the removal of organics in an existing dithc that does not have return sludge and which is not preceded by primary sedimentation. To accomplish this objective, samples of the wastewater influent and effluent were collected from the wastewater treatment plant of Mansfield, Texas during practical training. These samples were collected over an extended period of time to obtain samples at various treatment temperatures. Two analytical tests, COD and suspended solids, were used to monitor the operation of the plant. The results show that high removal efficiencies were obtained at high temperatures, with the efficiency decreasing as the temperature decreased to approximatedly 16$^{\circ}C$, at which point the efficienty tmeperature relationship appeared to stabilize.

• Journal of Information Processing Systems
• /
• v.17 no.6
• /
• pp.1057-1070
• /
• 2021

Case-related news filtering is crucial in legal text mining and divides news into case-related and case-unrelated categories. Because case-related news originates from various fields and has different writing styles, it is difficult to establish complete filtering rules or keywords for data collection. In addition, the labeled corpus for case-related news is sparse; therefore, to train a high-performance classification model, it is necessary to annotate the corpus. To address this challenge, we propose topic-enhanced positive-unlabeled learning, which selects positive and negative samples guided by topics. Specifically, a topic model based on a variational autoencoder (VAE) is trained to extract topics from unlabeled samples. By using these topics in the iterative process of positive-unlabeled (PU) learning, the accuracy of identifying case-related news can be improved. From the experimental results, it can be observed that the F1 value of our method on the test set is 1.8% higher than that of the PU learning baseline model. In addition, our method is more robust with low initial samples and high iterations, and compared with advanced PU learning baselines such as nnPU and I-PU, we obtain a 1.1% higher F1 value, which indicates that our method can effectively identify case-related news.