• Speech Sciences
• /
• v.10 no.4
• /
• pp.149-157
• /
• 2003

In this paper, we proposed the principal component analysis (PCA) fuzzy mixture model for speaker identification. A PCA fuzzy mixture model is derived from the combination of the PCA and the fuzzy version of mixture model with diagonal covariance matrices. In this method, the feature vectors are first transformed by each speaker's PCA transformation matrix to reduce the correlation among the elements. Then, the fuzzy mixture model for speaker is obtained from these transformed feature vectors with reduced dimensions. The orthogonal Gaussian Mixture Model (GMM) can be derived as a special case of PCA fuzzy mixture model. In our experiments, with having the number of mixtures equal, the proposed method requires less training time and less storage as well as shows better speaker identification rate compared to the conventional GMM. Also, the proposed one shows equal or better identification performance than the orthogonal GMM does.

• Journal of Korea Multimedia Society
• /
• v.7 no.2
• /
• pp.156-163
• /
• 2004

In this paper, we propose an adaptive face region detection and real-time face identification algorithm using face feature evaluation function. The proposed algorithm can detect exact face region adaptively by using skin color information for races as well as intensity and elliptical masking method. And also, it improves face recognition efficiency using geometrical face feature and geometric evaluation function between features. The proposed algorithm can be used for the development of biometric and security system areas. In the experiment, the superiority of the proposed method has been tested using real image, the proposed algorithm shows more improved recognition efficiency as well as face region detection efficiency than conventional method.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.22 no.2
• /
• pp.168-173
• /
• 2012

This paper presents a new human identification algorithm using partial features of the uncovered portion of face when a person wears a mask. After the face area is detected, the feature is extracted from the eye area above the mask. The identification process is performed by comparing the acquired one with the registered features. For extracting features SIFT(scale invariant feature transform) algorithm is used. The extracted features are independent of brightness and size- and rotation-invariant for the image. The experiment results show the effectiveness of the suggested algorithm.

• Journal of Information Processing Systems
• /
• v.19 no.5
• /
• pp.673-687
• /
• 2023

Convolutional neural network-based deep learning technology is the most commonly used in image identification, but it requires large-scale data for training. Therefore, application in specific fields in which data acquisition is limited, such as in the military, may be challenging. In particular, the identification of ground weapon systems is a very important mission, and high identification accuracy is required. Accordingly, various studies have been conducted to achieve high performance using small-scale data. Among them, the ensemble method, which achieves excellent performance through the prediction average of the pre-trained models, is the most representative method; however, it requires considerable time and effort to find the optimal combination of ensemble models. In addition, there is a performance limitation in the prediction results obtained by using an ensemble method. Furthermore, it is difficult to obtain the ensemble effect using models with imbalanced classification accuracies. In this paper, we propose a transfer learning-based feature fusion technique for heterogeneous models that extracts and fuses features of pre-trained heterogeneous models and finally, fine-tunes hyperparameters of the fully connected layer to improve the classification accuracy. The experimental results of this study indicate that it is possible to overcome the limitations of the existing ensemble methods by improving the classification accuracy through feature fusion between heterogeneous models based on transfer learning.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.49 no.2
• /
• pp.214-223
• /
• 2016

Joint time-frequency images of the broadband acoustic echoes of six fish species were obtained using the smoothed pseudo-Wigner-Ville distribution (SPWVD). The acoustic features were extracted by changing the sliced window widths and dividing the time window by a 0.02-ms interval and the frequency window by a 20-kHz bandwidth. The 22 spectrum amplitudes obtained in the time and frequency domains of the SPWVD images were fed as input parameters into an artificial neural network (ANN) to verify the effectiveness for species-dependent features related to fish species identification. The results showed that the time-frequency approach improves the extraction of species-specific features for species identification from broadband echoes, compare with time-only or frequency-only features. The ANN classifier based on these acoustic feature components was correct in approximately 74.5% of the test cases. In the future, the identification rate will be improved using time-frequency images with reduced dimensions of the broadband acoustic echoes as input for the ANN classifier.

• The Journal of the Acoustical Society of Korea
• /
• v.22 no.7
• /
• pp.519-527
• /
• 2003

Speech is much influenced by the existence of outliers which are introduced by such an unexpected happenings as additive background noise, change of speaker's utterance pattern and voice detection errors. These kinds of outliers may result in severe degradation of speaker recognition performance. In this paper, we proposed the GMM based on robust principal component analysis (RPCA-GMM) using M-estimation to solve the problems of both ouliers and high dimensionality of training feature vectors in speaker identification. Firstly, a new feature vector with reduced dimension is obtained by robust PCA obtained from M-estimation. The robust PCA transforms the original dimensional feature vector onto the reduced dimensional linear subspace that is spanned by the leading eigenvectors of the covariance matrix of feature vector. Secondly, the GMM with diagonal covariance matrix is obtained from these transformed feature vectors. We peformed speaker identification experiments to show the effectiveness of the proposed method. We compared the proposed method (RPCA-GMM) with transformed feature vectors to the PCA and the conventional GMM with diagonal matrix. Whenever the portion of outliers increases by every 2%, the proposed method maintains almost same speaker identification rate with 0.03% of little degradation, while the conventional GMM and the PCA shows much degradation of that by 0.65% and 0.55%, respectively This means that our method is more robust to the existence of outlier.

• Journal of Korea Multimedia Society
• /
• v.14 no.1
• /
• pp.1-14
• /
• 2011

While the circulation of images become active, various requirements to manage increasing database are raised. The content-based technology is one of methods to satisfy these requirements. The image is represented by feature vectors extracted by various methods in the content-based technology. The global feature method insures fast matching speed because the feature vector extracted by the global feature method is formed into a standard shape. The global feature extraction methods are classified into two categories, the spatial feature extraction and statistical feature extraction. And each group is divided by what kind of information is used, color feature or gray scale feature. In this paper, we introduce various global feature extraction technologies and compare their performance by accuracy, recall-precision graph, ANMRR, feature vector size and matching time. According to the experiments, the spatial features show good performance in non-geometrical modifications, and the extraction technologies that use color and histogram feature show the best performance.

• Journal of Korea Multimedia Society
• /
• v.23 no.8
• /
• pp.927-939
• /
• 2020

Gait recognition can identify people's identity from a long distance, which is very important for improving the intelligence of the monitoring system. Among many human features, gait features have the advantages of being remotely available, robust, and secure. Traditional gait feature extraction, affected by the development of behavior recognition, can only rely on manual feature extraction, which cannot meet the needs of fine gait recognition. The emergence of deep convolutional neural networks has made researchers get rid of complex feature design engineering, and can automatically learn available features through data, which has been widely used. In this paper,conduct feature metric learning in the three-dimensional space by combining the three-dimensional convolution features of the gait sequence and the Siamese structure. This method can capture the information of spatial dimension and time dimension from the continuous periodic gait sequence, and further improve the accuracy and practicability of gait recognition.

• Journal of the Korean Society of Industry Convergence
• /
• v.24 no.1
• /
• pp.1-7
• /
• 2021

In this paper, we presented the experimental basis for the theoretical background and robustness of the Convolutional Neural Network for object recognition based on artificial intelligence. An experimental result was performed to visualize the weighting filters and feature maps for each layer to determine what characteristics CNN is automatically generating. experimental results were presented on the trend of learning error and identification error rate by checking the relevance of the weight filter and feature map for learning error and identification error. The weighting filter and characteristic map are presented as experimental results. The automatically generated characteristic quantities presented the results of error rates for moving and rotating robustness to geometric changes.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.5
• /
• pp.1755-1777
• /
• 2022

The development of wireless communication technology has led to the underutilization of radio spectra. To address this limitation, an intelligent cognitive radio network was developed. Specific emitter identification (SEI) is a key technology in this network. However, in realistic non-cooperative scenarios, the system may detect signal classes beyond those in the training database, and only a few labeled signal samples are available for network training, both of which deteriorate identification performance. To overcome these challenges, a meta-learning-based open-set identification system is proposed for SEI. First, the received signals were pre-processed using bi-spectral analysis and a Radon transform to obtain signal representation vectors, which were then fed into an open-set SEI network. This network consisted of a deep feature extractor and an intrinsic feature memorizer that can detect signals of unknown classes and classify signals of different known classes. The training loss functions and the procedures of the open-set SEI network were then designed for parameter optimization. Considering the few-shot problems of open-set SEI, meta-training loss functions and meta-training procedures that require only a few labeled signal samples were further developed for open-set SEI network training. The experimental results demonstrate that this approach outperforms other state-of-the-art SEI methods in open-set scenarios. In addition, excellent open-set SEI performance was achieved using at least 50 training signal samples, and effective operation in low signal-to-noise ratio (SNR) environments was demonstrated.