Takehisa Yairi

Coupled Nonnegative Matrix Factorization Unmixing for Hyperspectral and Multispectral Data Fusion

Coupled nonnegative matrix factorization (CNMF) unmixing is proposed for the fusion of low-spatial-resolution hyperspectral and high-spatial-resolution multispectral data to produce fused data with high spatial and spectral resolutions. Both hyperspectral and multispectral data are alternately unmixed into end member and abundance matrices by the CNMF algorithm based on a linear spectral mixture model. Sensor observation models that relate the two data are built into the initialization matrix of each NMF unmixing procedure. This algorithm is physically straightforward and easy to implement owing to its simple update rules. Simulations with various image data sets demonstrate that the CNMF algorithm can produce high-quality fused data both in terms of spatial and spectral domains, which contributes to the accurate identification and classification of materials observed at a high spatial resolution.

Hyperspectral, multispectral, and panchromatic data fusion based on coupled non-negative matrix factorization

Coupled non-negative matrix factorization (CNMF) is applied to hyperspectral, multispectral, and panchromatic data fusion. This unmixing based method extracts and fuses hyperspectral endmember spectra and high-spatial-resolution abundance maps using these three data. An experiment with the synthetic data simulating ALOS-3 (advanced land observing satellite 3) dataset shows that the CNMF method has a possibility to produce fused data which have both high spatial and spectral resolutions with smaller spectral distortion.

Anomaly Detection for Autonomous Inspection of Space Facilities using Camera Images

For the purpose of realizing autonomous inspection of space facilities, this paper addresses the problem of anomaly detection from real-world images captured by free-flying space cameras. To cope with computer vision problems in space, we apply view-based and patch-based approach to represent features of image, and one-class SVM is used for classification. Anomaly detection framework is shown, which deals with a large amount of image patches fast and economically. The experimental result applied to anomaly detection from the Space Shuttle tile image was also shown, and it demonstrates favorable performance for anomaly detection problems

Autonomous generation of reflexion-based robot controller using inductive learning

Abstract The paper proposes a novel architecture for autonomously generating and managing a robot control system, aiming for the application to planetary rovers which will move in a partially unknown, unstructured environment. The proposed architecture is similar to the well known subsumption architecture in that the movements are governed by a network of various reflexion patterns. The major departures are that firstly it utilizes inductive learning to automatically generate and modify a control architecture, which is, if human is to do, quite a difficult and time consuming task, secondly it employs the concept of “goal sensor” to deal with the system goal more explicitly, and thirdly it compiles the planning results into a reflexion network and decision trees to maintain the strong features of reflexion based planner such as real-timeness, robustness and extensibility. The architecture has been applied to movement control of a certain rover in computer simulations and simple experiments, in which its effectiveness and characteristics have been cleared.

Coupled non-negative matrix factorization (CNMF) for hyperspectral and multispectral data fusion: Application to pasture classification

Coupled non-negative matrix factorization (CNMF) is introduced for hyperspectral and multispectral data fusion. The CNMF fused data have little spectral distortion while enhancing spatial resolution of all hyperspectral band images owing to its unmixing based algorithm. CNMF is applied to the synthetic dataset generated from real airborne hyperspectral data taken over pasture area. The spectral quality of fused data is evaluated by the classification accuracy of pasture types. The experiment result shows that CNMF enables accurate identification and classification of observed materials at fine spatial resolution.

A comparison study of feature spaces and classification methods for facial expression recognition

Facial expression recognition (FER) is important for robots and computers to achieve natural interaction with human. Over the years, researchers have proposed different feature descriptors, implemented different classification methods, and carried out test experiments on different datasets in realizing an automatic FER system. While achieving good performance, the most efficient feature space and classification method for FER remain unknown due to lack of comparison study. We performed comprehensive comparison experiments with five popular feature spaces in computer vision field and seven classification methods with four unique facial expression datasets. Our contributions in this work includes: (1) identified most efficient feature space for FER, (2) investigated effect of image resolutions on FER performances, and (3) obtained best FER performance by using AdaBoost algorithm for feature selection and Support Vector Machine for image classification.

Anomaly detection from multivariate time-series with sparse representation.

Anomaly detection from sensor data is an important data mining application for efficient and secure operation of complicated systems. In this study, we propose a novel anomaly detection method for multivariate time-series to capture relationships of variables and time-domain correlations simultaneously, without assuming any generative models of signals. The supposed framework in this study is a semi-supervised anomaly detection where we seek unusual parts of test data compared with reference data. The proposed method is based on feature extraction with sparse representation and relationship learning with dimensionality reduction. Our idea comes from the similarity between a sparse feature matrix extracted from multivariate time-series and a term-document matrix. We conducted experiments with synthetic and simulated data, and confirmed that the proposed method successfully detected anomalies in multivariate time-series signals. Especially, it demonstrated superior performance with anomalies in which only relationships of time-series patterns are changed from reference data (multivariate anomalies).

Generalized BRIEF: A Novel Fast Feature Extraction Method for Robust Hand Detection

Most literatures have been relying on image processing approaches such as skin detection and depth thresholding for hand detection. These techniques are restricted by strong assumptions and normally possess low robustness in actual applications. In this paper, we focus on an appearance approach and propose a new feature extraction method based on sparse pixel-pair wise intensity comparisons for hand detection. Our method can be viewed as a generalized BRIEF descriptor and can be easily adopted for other object detection or recognition tasks. We perform extensive experiments and prove that our method achieves comparable results with normal, noisy, and occluded hand images in term of both test accuracy and ROC. The main contributions of our work are threefold: 1) We introduce a new and simple feature extraction method that is robust against image noise, cluttered backgrounds, and partial occlusion. 2) Combined with AdaBoost, we show that the new feature descriptor is effective for hand detection. 3) The new feature descriptor has been rigorously compared with existing feature descriptors with a new hand database that has very challenging image backgrounds.

Identification of Causal Variables for Building Energy Fault Detection by Semi-supervised LDA and Decision Boundary Analysis

This paper addresses the identification problem of causal variables for the system anomaly. In real-world complicated systems, even experts often fail to specify causal factors, thus they attempt to detect the anomaly with exploratory heuristics. Our goal is to offer further information that supports anomaly cause analysis using the incomplete empirical knowledge. Proposed technique discovers responsible factors for the fault by leveraging domain knowledge with an effective combination of semi-supervised linear discriminant analysis (LDA) and boundary-based discriminative subspace identification method. Experimental results on synthetic and real dataset confirmed validity of our approach. Moreover, we applied this method to the building energy fault diagnosis and succeeded in extracting causal variables for energy waste in a building.

Integrating Data Mining Techniques and Design Information Management for Failure Prevention

Stories of the recent failures in complex systems tell us that they could have been avoided if the right information was presented to the right person at the right time. We propose a method for fault detection of spacecrafts by mining association rules from house keeping data. We also argue that merely detecting anomalies is not enough for failure prevention. We present a framework of design information management in order to capture and use design rationale for failure prevention. We believe that the framework provides the basis for improved development process and effective anomaly handling.