NEC (Japan)

This paper reports on an optimum dynamic progxamming (DP) based time-normalization algorithm for spoken word recognition. First, a general principle of time-normalization is given using time-warping function. Then, two time-normalized distance definitions, called symmetric and asymmetric forms, are derived from the principle. These two forms are compared with each other through theoretical discussions and experimental studies. The symmetric form algorithm superiority is established. A new technique, called slope constraint, is successfully introduced, in which the warping function slope is restricted so as to improve discrimination between words in different categories. The effective slope constraint characteristic is qualitatively analyzed, and the optimum slope constraint condition is determined through experiments. The optimized algorithm is then extensively subjected to experimental comparison with various DP-algorithms, previously applied to spoken word recognition by different research groups. The experiment shows that the present algorithm gives no more than about two-thirds errors, even compared to the best conventional algorithm.

On the basis of realistic tight-binding band-structure calculations, we predict that carbon microtubules exhibit striking variations in electronic transport, from metallic to semiconducting with narrow and moderate band gaps, depending on the diameter of the tubule and on the degree of helical arrangement of the carbon hexagons. The origin of this drastic variation in the band structure is explained in terms of the two-dimensional band structure of graphite.

Influences on the semiconductor laser properties of external optical feedback, i.e., return of a portion of the laser output from a reflector external to the laser cavity, have been examined. Experimental observations with a single mode laser is presented with analysis based on a compound cavity laser model, which has been found to explain essential features of the experimental results. In particular, it has been demonstrated that a laser with external feedback can be multistable and show hysteresis phenomena, analogous to those of non-linear Fabry-Perot resonator. It has also been shown that the dynamic properties of injection lasers are significantly affected by external feedback, depending on interference conditions between returned light and the field inside the laser diode.

Textural features corresponding to human visual perception are very useful for optimum feature selection and texture analyzer design. We approximated in computational form six basic textural features, namely, coarseness, contrast, directionality, line-likeness, regularity, and roughness. In comparison with psychological measurements for human subjects, the computational measures gave good correspondences in rank correlation of 16 typical texture patterns. Similarity measurements using these features were attempted. The discrepancies between human vision and computerized techniques that we encountered in this study indicate fundamental problems in digital analysis of textures. Some of them could be overcome by analyzing their causes and using more sophisticated techniques.

A Python package for working with the Information Bottleneck [Tishby, Pereira, Bialek 2001] and the Deterministic (and Generalized) Information Bottleneck [Strouse and Schwab 2016]. Embo is especially geared towards the analysis of concrete, finite-size data sets. See on PyPI <strong>How to cite:</strong> Piasini, E., Filipowicz, A.L.S., Levine, J. and Gold, J.I., 2021. Embo: a Python package for empirical data analysis using the Information Bottleneck. <em>Journal of Open Research Software</em>, 9(1), p.10. DOI: http://doi.org/10.5334/jors.322

This paper presents an overview of color and texture descriptors that have been approved for the Final Committee Draft of the MPEG-7 standard. The color and texture descriptors that are described in this paper have undergone extensive evaluation and development during the past two years. Evaluation criteria include effectiveness of the descriptors in similarity retrieval, as well as extraction, storage, and representation complexities. The color descriptors in the standard include a histogram descriptor that is coded using the Haar transform, a color structure histogram, a dominant color descriptor, and a color layout descriptor. The three texture descriptors include one that characterizes homogeneous texture regions and another that represents the local edge distribution. A compact descriptor that facilitates texture browsing is also defined. Each of the descriptors is explained in detail by their semantics, extraction and usage. The effectiveness is documented by experimental results.

Abstract A nanometre-scale superconducting electrode connected to a reservoir via a Josephson junction constitutes an artificial two-level electronic system: a single-Cooper-pair box. The two levels consist of charge states (differing by 2e, where e is the electronic charge) that are coupled by tunnelling of Cooper pairs through the junction. Although the two-level system is macroscopic, containing a large number of electrons, the two charge states can be coherently superposed1, 2, 3, 4. The Cooper-pair box has therefore been suggested5, 6, 7 as a candidate for a quantum bit or ‘qubit’—the basic component of a quantum computer. Here we report the observation of quantum oscillations in a single-Cooper-pair box. By applying a short voltage pulse via a gate electrode, we can control the coherent quantum state evolution: the pulse modifies the energies of the two charge states non-adiabatically, bringing them into resonance. The resulting state—a superposition of the two charge states—is detected by a tunnelling current through a probe junction. Our results demonstrate electrical coherent control of a qubit in a solid-state electronic device.

We report high resolution electron microscope (HREM) observations and atomistic simulations of the bending of single and multi-walled carbon nanotubes under mechanical duress. Single and multiple kinks are observed at high bending angles. Their occurrence is quantitatively explained by the simulations, which use a realistic many-body potential for the carbon atoms. We show that the bending is fully reversible up to very large bending angles, despite the occurrence of kinks and highly strained tube regions. This is due to the remarkable flexibility of the hexagonal network, which resists bond breaking and bond switching up to very high strain values.

We have observed coherent time evolution between two quantum states of a superconducting flux qubit comprising three Josephson junctions in a loop. The superposition of the two states carrying opposite macroscopic persistent currents is manipulated by resonant microwave pulses. Readout by means of switching-event measurement with an attached superconducting quantum interference device revealed quantum-state oscillations with high fidelity. Under strong microwave driving, it was possible to induce hundreds of coherent oscillations. Pulsed operations on this first sample yielded a relaxation time of 900 nanoseconds and a free-induction dephasing time of 20 nanoseconds. These results are promising for future solid-state quantum computing.

A secure communication network with quantum key distribution in a metropolitan area is reported. Six different QKD systems are integrated into a mesh-type network. GHz-clocked QKD links enable us to demonstrate the world-first secure TV conferencing over a distance of 45km. The network includes a commercial QKD product for long-term stable operation, and application interface to secure mobile phones. Detection of an eavesdropper, rerouting into a secure path, and key relay via trusted nodes are demonstrated in this network.

We present our experiences to date building ONOS (Open Network Operating System), an experimental distributed SDN control platform motivated by the performance, scalability, and availability requirements of large operator networks. We describe and evaluate two ONOS prototypes. The first version implemented core features: a distributed, but logically centralized, global network view; scale-out; and fault tolerance. The second version focused on improving performance. Based on experience with these prototypes, we identify additional steps that will be required for ONOS to support use cases such as core network traffic engineering and scheduling, and to become a usable open source, distributed network OS platform that the SDN community can build upon.

We present microscopic total-energy calculations which provide a cohesive property and electronic structures of a new form of solid carbon, the face-centered-cubic ${\mathrm{C}}_{60}$ crystal (fcc ${\mathrm{C}}_{60}$). We find that ${\mathrm{C}}_{60}$ clusters are condensed by van der Waals force, and that the resulting fcc ${\mathrm{O}}_{60}$ is a novel semiconductor with direct energy gap of 1.5 eV at the Brillouin-zone boundary (X point). We also find that an ``impurity'' ${\mathrm{C}}_{60}$K cluster induces a shallow donor state in fcc ${\mathrm{C}}_{60}$.

Thick GaN layers were grown by hydride vapor phase epitaxy (HVPE) with the aim of using these layers as a homoepitaxial substrate to improve device quality of laser diodes or light emitting diodes. HVPE is very useful for thick layer growth since the growth rate can reach from several ten up to one hundred micron per hour. In this experiment, the growth began as selective growth through openings formed in a SiO 2 mask. Facets consisting of {1101} planes were formed in the early stage and a continuous film developed from the coalescence of these facets on the SiO 2 mask. As a result, GaN layers with a dislocation density as low as 6×10 7 cm -2 were grown on 2-inch-diameter sapphire wafers. These GaN layers were crack-free and had mirror-like surface.

Extraordinary angle-sensitive light propagation, which we call a superprism phenomenon, was demonstrated at optical wavelength in photonic crystals with three-dimensional-periodic structure fabricated on Si substrate. The propagation beam was swung from $\ensuremath{-}90\ifmmode^\circ\else\textdegree\fi{}$ to $+90\ifmmode^\circ\else\textdegree\fi{}$ with a slight change in the incident angle within $\ifmmode\pm\else\textpm\fi{}12\ifmmode^\circ\else\textdegree\fi{}.$ This effect together with wavelength sensitivity is at least two orders of magnitude stronger than that of the conventional prism. The incident-angle dependence including negative refraction and multiple beam branching was interpreted from highly anisotropic dispersion surfaces derived by photonic band calculation. These phenomena will be available to fabricate microscale light circuits on Si with LSI-compatible lithography techniques.

The wetting and capillarity of carbon nanotubes were studied in detail here. Nanotubes are not "super-straws," although they can be wet and filled by substances having low surface tension, such as sulfur, selenium, and cesium, with an upper limit to this tension less than 200 millinewtons per meter. This limit implies that typical pure metals will not be drawn into the inner cavity of nanotubes through capillarity, whereas water and organic solvents will. These results have important implications for the further use of carbon nanotubes in experiments on a nanometer scale.

A self-recovering equalization algorithm, which is employed in multilevel amplitude-modulated data transmission, is presented. Such a self-recovering equalizer has been required when time-division multiplexed (TDM) voice or picturephone PCM signals must be transmitted over the existing frequency-division multiplexed (FDM) transmission channel. The present self-recovering equalizer is quite simple, as is a conventional binary equalizer. The convergence processes of the present self-recovering equalizer are shown by computer simulation. Some theoretical considerations on this convergence process are also added.

In order to accelerate an algorithm for test generation, it is necessary to reduce the number of backtracks in the algorithm and to shorten the process time between backtracks. In this paper, we consider several techniques to accelerate test generation and present a new test generation algorithm called FAN (fan-out-oriented test generation algorithm). It is shown that the FAN algorithm is faster and more efficient than the PODEM algorithm reported by Goel. We also present an automatic test generation system composed of the FAN algorithm and the concurrent fault simulation. Experimental results on large combinational circuits of up to 3000 gates demonstrate that the system performs test generation very fast and effectively.

This paper presents a novel large-scale dataset and comprehensive baselines for end-to-end pedestrian detection and person recognition in raw video frames. Our baselines address three issues: the performance of various combinations of detectors and recognizers, mechanisms for pedestrian detection to help improve overall re-identification (re-ID) accuracy and assessing the effectiveness of different detectors for re-ID. We make three distinct contributions. First, a new dataset, PRW, is introduced to evaluate Person Re-identification in the Wild, using videos acquired through six synchronized cameras. It contains 932 identities and 11,816 frames in which pedestrians are annotated with their bounding box positions and identities. Extensive benchmarking results are presented on this dataset. Second, we show that pedestrian detection aids re-ID through two simple yet effective improvements: a cascaded fine-tuning strategy that trains a detection model first and then the classification model, and a Confidence Weighted Similarity (CWS) metric that incorporates detection scores into similarity measurement. Third, we derive insights in evaluating detector performance for the particular scenario of accurate person re-ID.

HUPO initiated the Plasma Proteome Project (PPP) in 2002. Its pilot phase has (1) evaluated advantages and limitations of many depletion, fractionation, and MS technology platforms; (2) compared PPP reference specimens of human serum and EDTA, heparin, and citrate-anti-coagulated plasma; and (3) created a publicly-available knowledge base (www.bioinformatics.med.umich.edu/hupo/ppp; www.ebi.ac.uk/pride). Thirty-five participating laboratories in 13 countries submitted datasets. Working groups addressed (a) specimen stability and protein concentrations; (b) protein identifications from 18 MS/MS datasets; (c) independent analyses from raw MS-MS spectra; (d) search engine performance, subproteome analyses, and biological insights; (e) antibody arrays; and (f) direct MS/SELDI analyses. MS-MS datasets had 15 710 different International Protein Index (IPI) protein IDs; our integration algorithm applied to multiple matches of peptide sequences yielded 9504 IPI proteins identified with one or more peptides and 3020 proteins identified with two or more peptides (the Core Dataset). These proteins have been characterized with Gene Ontology, InterPro, Novartis Atlas, OMIM, and immunoassay-based concentration determinations. The database permits examination of many other subsets, such as 1274 proteins identified with three or more peptides. Reverse protein to DNA matching identified proteins for 118 previously unidentified ORFs. We recommend use of plasma instead of serum, with EDTA (or citrate) for anticoagulation. To improve resolution, sensitivity and reproducibility of peptide identifications and protein matches, we recommend combinations of depletion, fractionation, and MS/MS technologies, with explicit criteria for evaluation of spectra, use of search algorithms, and integration of homologous protein matches. This Special Issue of PROTEOMICS presents papers integral to the collaborative analysis plus many reports of supplementary work on various aspects of the PPP workplan. These PPP results on complexity, dynamic range, incomplete sampling, false-positive matches, and integration of diverse datasets for plasma and serum proteins lay a foundation for development and validation of circulating protein biomarkers in health and disease.

Nowadays, multivariate time series data are increasingly collected in various real world systems, e.g., power plants, wearable devices, etc. Anomaly detection and diagnosis in multivariate time series refer to identifying abnormal status in certain time steps and pinpointing the root causes. Building such a system, however, is challenging since it not only requires to capture the temporal dependency in each time series, but also need encode the inter-correlations between different pairs of time series. In addition, the system should be robust to noise and provide operators with different levels of anomaly scores based upon the severity of different incidents. Despite the fact that a number of unsupervised anomaly detection algorithms have been developed, few of them can jointly address these challenges. In this paper, we propose a Multi-Scale Convolutional Recurrent Encoder-Decoder (MSCRED), to perform anomaly detection and diagnosis in multivariate time series data. Specifically, MSCRED first constructs multi-scale (resolution) signature matrices to characterize multiple levels of the system statuses in different time steps. Subsequently, given the signature matrices, a convolutional encoder is employed to encode the inter-sensor (time series) correlations and an attention based Convolutional Long-Short Term Memory (ConvLSTM) network is developed to capture the temporal patterns. Finally, based upon the feature maps which encode the inter-sensor correlations and temporal information, a convolutional decoder is used to reconstruct the input signature matrices and the residual signature matrices are further utilized to detect and diagnose anomalies. Extensive empirical studies based on a synthetic dataset and a real power plant dataset demonstrate that MSCRED can outperform state-ofthe-art baseline methods.