• Journal of KIISE:Software and Applications
• /
• v.36 no.11
• /
• pp.960-966
• /
• 2009

A manifold is used to represent a relationship between high-dimensional data samples in low-dimensional space. In human pose estimation, it is created in low-dimensional space for processing image and 3D body configuration data. Manifold learning is to build a manifold. But it is vulnerable to silhouette variations. Such silhouette variations are occurred due to view-change, person-change, distance-change, and noises. Representing silhouette variations in a single manifold is impossible. In this paper, we focus a silhouette variation problem occurred by view-change. In previous view invariant pose estimation methods based on manifold learning, there were two ways. One is modeling manifolds for all view points. The other is to extract view factors from mapping functions. But these methods do not support one by one mapping for silhouettes and corresponding body configurations because of unsupervised learning. Modeling manifold and extracting view factors are very complex. So we propose a method based on triple manifolds. These are view manifold, pose manifold, and body configuration manifold. In order to build manifolds, we employ biased manifold learning. After building manifolds, we learn mapping functions among spaces (2D image space, pose manifold space, view manifold space, body configuration manifold space, 3D body configuration space). In our experiments, we could estimate various body poses from 24 view points.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.9
• /
• pp.4317-4335
• /
• 2018

Semi-supervised multi-view latent space learning is gaining considerable popularity recently in many machine learning applications due to the high cost and difficulty to obtain the large amount of label information of data. Although some semi-supervised multi-view latent space learning methods have been presented, there is still much space for improvement: 1) How to learn latent discriminant intact feature representations by employing data of multiple views; 2) How to exploit the manifold structure of both labeled and unlabeled point in the learned latent intact space effectively. To address the above issues, we propose an approach called semi-supervised multi-view manifold discriminant intact space learning ($SM^2DIS$) for image classification in this paper. $SM^2DIS$ aims to seek a manifold discriminant intact space for data of different views by making use of both the discriminant information of labeled data and the manifold structure of both labeled and unlabeled data. Experimental results on MNIST, COIL-20, Multi-PIE, and Caltech-101 databases demonstrate the effectiveness and robustness of our proposed approach.

• Korean Journal of Remote Sensing
• /
• v.26 no.6
• /
• pp.693-703
• /
• 2010

High spectral resolution of hyperspectral data enables analysis of complex natural phenomena that is reflected on the data nonlinearly. Although many manifold learning methods have been developed for such problems, most methods do not consider the spatial correlation between samples that is inherent and useful in remote sensing data. We propose a manifold learning method which directly combines the spatial proximity and the spectral similarity through kernel PCA framework. A gain factor caused by spatial proximity is first modelled with a heat kernel, and is added to the original similarity computed from the spectral values of a pair of samples. Parameters are tuned with intelligent grid search (IGS) method for the derived manifold coordinates to achieve optimal classification accuracies. Of particular interest is its performance with small training size, because labelled samples are usually scarce due to its high acquisition cost. The proposed spatial kernel PCA (KPCA) is compared with PCA in terms of classification accuracy with the nearest-neighbourhood classification method.

• Journal of Electrical Engineering and Technology
• /
• v.13 no.5
• /
• pp.2099-2106
• /
• 2018

This study presents a novel approach of discriminative feature vectors based on manifold learning using nonlinear dimension reduction (DR) technique to improve loss function, and combine with the Adversarial examples to regularize the object function for image classification. The traditional convolutional neural networks (CNN) with many new regularization approach has been successfully used for image classification tasks, and it achieved good results, hence it costs a lot of Calculated spacing and timing. Significantly, distrinct from traditional CNN, we discriminate the feature vectors for objects without empirically-tuned parameter, these Discriminative features intend to remain the lower-dimensional relationship corresponding high-dimension manifold after projecting the image feature vectors from high-dimension to lower-dimension, and we optimize the constrains of the preserving local features based on manifold, which narrow the mapped feature information from the same class and push different class away. Using Adversarial examples, improved loss function with additional regularization term intends to boost the Robustness and generalization of neural network. experimental results indicate that the approach based on discriminative feature of manifold learning is not only valid, but also more efficient in image classification tasks. Furthermore, the proposed approach achieves competitive classification performances for three benchmark datasets : MNIST, CIFAR-10, SVHN.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2008.04a
• /
• pp.135-138
• /
• 2008

최근 인간을 모방하는 휴머노이드 로봇(Humanoid robot)에 대한 관심이 증가함에 따라, 기계공학, 생체공학, 제어이론 등 여러 분야에서 관련 연구가 활발히 진행되고 있다. 이에 본 논문에서는 액츄에이터(Actuator)가 없이 경사진 지면을 걸을 수 있는 두 발을 가진 패시브 로봇(Passive robot)을 대상으로 강화학습과 메니폴드(Manifold control) 기법을 사용하여 안정적으로 걸을 수 있도록 제어기(Controller)를 설계하는 방안을 고려한다.

• Korean Journal of Remote Sensing
• /
• v.29 no.5
• /
• pp.509-516
• /
• 2013

Since traditional linear feature extraction methods are unable to handle nonlinear characteristics often exhibited in hyperspectral imagery, nonlinear feature extraction, also known as manifold learning, is receiving increased attention in hyperspectral remote sensing society as well as other community. A most widely used manifold Isomap is generally promising good results in classification and spectral unmixing tasks, but significantly high computational overhead is problematic, especially for large scale remotely sensed data. A small subset of distinguishing points, referred to as landmarks, is proposed as a solution. This study proposes a new robust and controllable landmark selection method based on the maximum volume of the simplex spanned by landmarks. The experiments are conducted to compare classification accuracies with standard deviation according to sampling methods, the number of landmarks, and processing time. The proposed method could employ both classification accuracy and computational efficiency.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.1
• /
• pp.368-391
• /
• 2018

Face recognition (FR) with a single sample per person (SSPP) is common in real-world face recognition applications. In this scenario, it is hard to predict intra-class variations of query samples by gallery samples due to the lack of sufficient training samples. Inspired by the fact that similar faces have similar intra-class variations, we propose a virtual sample generating algorithm called k nearest neighbors based virtual sample generating (kNNVSG) to enrich intra-class variation information for training samples. Furthermore, in order to use the intra-class variation information of the virtual samples generated by kNNVSG algorithm, we propose image set based multimanifold discriminant learning (ISMMDL) algorithm. For ISMMDL algorithm, it learns a projection matrix for each manifold modeled by the local patches of the images of each class, which aims to minimize the margins of intra-manifold and maximize the margins of inter-manifold simultaneously in low-dimensional feature space. Finally, by comprehensively using kNNVSG and ISMMDL algorithms, we propose k nearest neighbor virtual image set based multimanifold discriminant learning (kNNMMDL) approach for single sample face recognition (SSFR) tasks. Experimental results on AR, Multi-PIE and LFW face datasets demonstrate that our approach has promising abilities for SSFR with expression, illumination and disguise variations.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.2
• /
• pp.148-155
• /
• 2014

In this paper a commute time embedding is implemented by organizing patches according to the graph-based metric, and its properties are investigated via changing the number of nodes on the graph.. It is shown that manifold embedding methods generate the intrinsic geometric structures when waveforms such as speech or music instrumental sound signals are embedded on the low dimensional Euclidean space. Basically manifold embedding algorithms only project the training samples on the graph into an embedding subspace but can not generalize the learning results to test samples. They are very effective for data clustering but are not appropriate for classification or recognition. In this paper a commute time guided transform is adopted to enhance the generalization ability and its performance is analyzed by applying it to the classification of 6 kinds of music instrumental sounds.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.22 no.2
• /
• pp.103-108
• /
• 2022

In this paper, we propose a new image analysis method based on regression manifold 3-D PCA. The proposed method is a new image analysis method consisting of a regression analysis algorithm with a structure designed based on an autoencoder capable of nonlinear expansion of manifold 3-D PCA and PCA for efficient dimension reduction when entering large-capacity image data. With the configuration of an autoencoder, a regression manifold 3-DPCA, which derives the best hyperplane through three-dimensional rotation of image pixel values, and a Bayesian rule structure similar to a deep learning structure, are applied. Experiments are performed to verify performance. The image is improved by utilizing the fine dust image, and accuracy performance evaluation is performed through the classification model. As a result, it can be confirmed that it is effective for deep learning performance.

• Journal of KIISE:Software and Applications
• /
• v.31 no.2
• /
• pp.182-188
• /
• 2004

This paper proposes to synthesize facial images from a few parameters for the pose and the expression of their constituent components. This parameterization makes the representation, storage, and transmission of face images effective. But it is difficult to parameterize facial images because variations of face images show a complicated nonlinear manifold in high-dimensional data space. To tackle this problem, we use an LLE (Locally Linear Embedding) technique for a good representation of face images, where the relationship among face images is preserving well and the projected manifold into the reduced feature space becomes smoother and more continuous. Next, we apply a snake model to estimate face feature values in the reduced feature space that corresponds to a specific pose and/or expression parameter. Finally, a synthetic face image is obtained from an interpolation of several neighboring face images in the vicinity of the estimated feature value. Experimental results show that the proposed method shows a negligible overlapping effect and creates an accurate and consistent synthetic face images with respect to changes of pose and/or expression parameters.