The articles in this month’s special issue provide insight into future computing technologies such as novel architectures, spintronic memories, and quantum computing.

The future of computing is at crossroads. The technological advances that have sustained the exponential growth of computing performance over the last several decades are slowing and the roadmap for future advances is uncertain. The phenomenal expansion of computing power has made computers ubiquitous, spawned a $300 billion semiconductor industry, enabled unprecedented global economic growth, and transformed many aspects of society at large. Emerging technologies are placing an ever-growing and changing demand on computing, especially the profusion of data from the Internet of Things, large-scale scientific experiments (high-energy physics, astronomy, and genomics), autonomous vehicles, social media (including video), national security systems, and the finance sector. Transmitting, storing, processing, and analyzing this data explosion with the requisite speed and performance—and enabling significant processing and analysis to occur locally or at network nodes (i.e., edge computing)—may require a radical departure from the traditional computing paradigm of von Neumann computing architectures running on CMOS-based digital logic. New paradigms will likely require a range of new devices, software, design and simulation tools, and benchmarking, and may ultimately require rethinking the tasks that computing machines undertake. Recently, government, industry, and academia collectively have recognized that the future of computing requires a new, multidisciplinary research and development agenda.

MOTIVATION: Protein-protein interactions are a key in virtually all biological processes. For a detailed understanding of the biological processes, the structure of the protein complex is essential. Given the current experimental techniques for structure determination, the vast majority of all protein complexes will never be solved by experimental techniques. In lack of experimental data, computational docking methods can be used to predict the structure of the protein complex. A common strategy is to generate many alternative docking solutions (atomic models) and then use a scoring function to select the best. The success of the computational docking technique is, to a large degree, dependent on the ability of the scoring function to accurately rank and score the many alternative docking models. RESULTS: Here, we present ProQDock, a scoring function that predicts the absolute quality of docking model measured by a novel protein docking quality score (DockQ). ProQDock uses support vector machines trained to predict the quality of protein docking models using features that can be calculated from the docking model itself. By combining different types of features describing both the protein-protein interface and the overall physical chemistry, it was possible to improve the correlation with DockQ from 0.25 for the best individual feature (electrostatic complementarity) to 0.49 for the final version of ProQDock. ProQDock performed better than the state-of-the-art methods ZRANK and ZRANK2 in terms of correlations, ranking and finding correct models on an independent test set. Finally, we also demonstrate that it is possible to combine ProQDock with ZRANK and ZRANK2 to improve performance even further. AVAILABILITY AND IMPLEMENTATION: http://bioinfo.ifm.liu.se/ProQDock CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

As noted on our webpage, the Proceedings of the IEEE specializes in surveys, reviews, and tutorials. The distinction between surveys and reviews is not always clear to readers and authors. Our working definition is that surveys discuss the current state of a subject and give an overview and summary of the published and ongoing work by researchers from all geographical regions and from all organizations contributing to the topic. Reviews contain more technical depth of the topics and usually discuss the technical background needed to understand and evaluate the contributions to the topics under investigation. Tutorials are teaching tools that present the background and current theory and methodologies of a topic. A tutorial is not about presenting new theory or methods, but may use current research as examples of applications of the techniques discussed. Currently, we publish many more surveys and reviews than tutorials.

Motivated by gas dynamics, we consider the heat diffusion equation which in its standard form allows propagation at an infinite speed, and is thus unphysical. Driven by phenomenological considerations rooted in analogies with RC-transmission lines, we introduce a modified 3D heat equation for an isotropic medium that admits MD internally passive circuit realization using resistances, nonlinear inductances, and capacitances of the type previously introduced elsewhere, which via standard techniques can be transformed into well-known Wave Digital networks in the discrete MD domain.

BACKGROUND: Mapping protein primary sequences to their three dimensional folds referred to as the 'second genetic code' remains an unsolved scientific problem. A crucial part of the problem concerns the geometrical specificity in side chain association leading to densely packed protein cores, a hallmark of correctly folded native structures. Thus, any model of packing within proteins should constitute an indispensable component of protein folding and design. RESULTS: In this study an attempt has been made to find, characterize and classify recurring patterns in the packing of side chain atoms within a protein which sustains its native fold. The interaction of side chain atoms within the protein core has been represented as a contact network based on the surface complementarity and overlap between associating side chain surfaces. Some network topologies definitely appear to be preferred and they have been termed 'packing motifs', analogous to super secondary structures in proteins. Study of the distribution of these motifs reveals the ubiquitous presence of typical smaller graphs, which appear to get linked or coalesce to give larger graphs, reminiscent of the nucleation-condensation model in protein folding. One such frequently occurring motif, also envisaged as the unit of clustering, the three residue clique was invariably found in regions of dense packing. Finally, topological measures based on surface contact networks appeared to be effective in discriminating sequences native to a specific fold amongst a set of decoys. CONCLUSIONS: Out of innumerable topological possibilities, only a finite number of specific packing motifs are actually realized in proteins. This small number of motifs could serve as a basis set in the construction of larger networks. Of these, the triplet clique exhibits distinct preference both in terms of composition and geometry.

Statistical modeling for content based retrieval is examined in the context of recent TREC Video benchmark exercise. The TREC Video exercise can be viewed as a test bed for evaluation and comparison of a variety of different algorithms on a set of high-level queries for multimedia retrieval. We report on the use of techniques adopted from statistical learning theory. Our method, as in most statistical methods, depend on training of models based on large data sets. A plethora of statistical models such the Gaussian mixture models, support vector machines etc. can be thought of, only a few of which are exploited in this preliminary report. Training requires a large amount of annotated (labeled) data. Thus, we explore use of active learning for the annotation engine that minimizes the number of training samples to be labeled for satisfactory performance.

Supporting semantic queries is a challenging problem in video retrieval. We propose the use of a lexicon of semantic concepts for handling the queries. We also propose automatic modeling of lexicon items using probabilistic techniques. We use Gaussian mixture models to build computational representations for a variety of semantic concepts including rocket-launch, outdoor greenery, sky etc. Training requires a large amount of annotated (labeled) data. Using the TREC Video test bed we compare the performance of this system supporting query by keywords with the conventional approach of query by example. Results demonstrate significant gains in performance using the automatically learnt models of semantic concepts.

We describe a system for content-based retrieval of video that involves a series of query interactions with the user. The proposed approach allows the user, iteratively and selectively, to integrate different feature- and model-based methods of querying in the search process. This allows the user to choose among different retrieved content, features and matching dimensions, and classifiers, as appropriate, given the query objective and interim retrieval results. We investigate several approaches for integrating featureand model-based queries and results in successive query rounds including iterative filtering, score aggregation, and relevance feedback searching. We describe experimental results of applying the interactive content-based retrieval method to an automatically indexed corpus of 11 hours of video.

We show that it is possible to have a large family of perfect reconstruction multidimensional (n-D) subband coding filter banks, which are causal, stable and IIR both at the analysis and at the synthesis ends. Complete parameterization of such classes of filters leading to design methods are given. Parameterizations in frequency domain terms as well as in terms of state space descriptions of filters are provided. The development is valid for two-band as well as for multi-band subband coding schemes. Our development is also valid for dilation matrices other than the conventional quincunx subsampling. Thus, sampling density ratios other than two are implicit in the development resulting from the fact that we consider multi-band subband coding schemes for arbitrary decimation and interpolation matrices.

We discuss several numerical methods for optimum feature selection for multivariate data based on maximum entropy and maximum likelihood criteria. Our point of view is to consider observed data x/sup 1/, x/sup 2/,..., x/sup N/ in R/sup d/ to be samples from some unknown pdf P. We project this data onto d directions, subsequently estimate the pdf of the univariate data, then find the maximum entropy (or likelihood) of all multivariate pdfs in R/sup d/ with marginals in these directions prescribed by the estimated univariate pdfs and finally maximize the entropy (or likelihood) further over the choice of these directions. This strategy for optimal feature selection depends on the method used to estimate univariate data.

We consider a parametric family of density functions of the type exp(-|x|/sup /spl alpha//2/) for modeling acoustic feature vectors used in automatic recognition of speech. The parameter /spl alpha/ is a measure of the impulsiveness as well as the nongaussian nature of the data. While previous work has focused on estimating the mean and the variance of the data here we attempt to estimate the impulsiveness /spl alpha/ from the data on a maximum likelihood basis. We show that there is a balance between /spl alpha/ and the number of data points N that must be satisfied before maximum likelihood estimation is carried out. Numerical experiments are performed on multidimensional vectors obtained from speech data.

In this paper we report recent improvements in voicemail transcription. Last year, the speaker independent and speaker adapted word error rates (WER) on the Voicemail Transcription task were reported at 41.94% and 38.18% respectively. This year, we report a relative improvement of 18% in the speaker independent performance and 11% in the speaker adapted performance over last year. This improvement is a result of some new algorithms and an increase in the amount of training data. In the following sections, we describe the contribution of several components to improving the word error rate. 1. INTRODUCTION In this paper we report recent improvements in voicemail transcription. The voicemail transcription task was introduced last year [1] as representing a style of conversational telephone speech that is somewhat different from the Switchboard and CallHome databases. Last year, the speaker independent and speaker adapted word error rates (WER) on this task were reported at 41.94% and 38...

Considers the problem of combining visual cues with audio signals for the purpose of improved automatic machine recognition of speech. Although significant progress has been made in the machine transcription of large-vocabulary continuous speech (LVCSR) over the last few years, the technology to date is most effective only under controlled conditions, such as low noise, speaker-dependent recognition, read speech (as opposed to conversational speech), etc. On the other hand, while augmenting the recognition of speech utterances with visual cues has attracted the attention of researchers over the last couple of years, most efforts in this domain can be considered to be only preliminary in the sense that, unlike LVCSR efforts, tasks have been limited to small vocabularies (e.g. commands, digits) and often to speaker-dependent training or isolated word speech, where word boundaries are artificially well-defined.

Complete parametrization of multiband linear phase biorthogonal filter banks are given. The method uses matrix reduction methods similar to the Hermite reduction method of linear system theory. Computational algorithms are derived for design, and examples are worked out.

We device a novel algorithm for the design of multiband perfect reconstruction (PR) linear phase FIR filter banks based on a technique akin to the Hermite reduction algorithm in system theory. The essential idea behind the algorithm stems from the fact that there is a large degree of freedom available in the multiband PR filter bank design. As a result, in an m-band problem it is almost always possible to specify n (n

Adaptive two-dimensional (2-D) fan filters based on least mean square algorithm and 2-D wave digital filter structure are presented. The weight updating equations as well as the equations for computing the gradient components of signals in two-dimensional wave digital filter adaptors are derived and specifically applied to wave digital quadrant and narrowband fan filters. Computer simulations are also conducted to show their potential applications to 2-D system identification and noise cancellation in noisy images.>

Following our recent work on the design and parametrization of one-dimensional, causal, stable, perfect reconstruction subband coding schemes, we show that very similar results hold in multi-dimensions as well. Such considerations are far more non-trivial, however, due to difficulties characteristic of the algebra of multi-variable polynomials and rational functions. As in 1-D, both frequency as well as state space domain descriptions of complete parametrization and design methods turn out to be feasible in the multidimensional case. While the need for causal, stable schemes in nonseparable subband coding is substantiated, at the same time we raise and discuss the issue of genericity of the PR property in multidimensional problems. Examples are included to partially illustrate the central theme of the work reported.>

We consider the multidimensional version of the problem of linear phase perfect reconstruction filter bank design. We give conditions for linear phase property of the filter bank, enumerate the number of symmetric and antisymmetric filters in the bank and give results on the possibility of construction of the entire filter bank if all filters in the bank except one are specified without any restriction other than the linear phase property.>

The performance of signal processing methods under model and environmental mismatch are discussed. The problem of signal detection in a random multipath channel is addressed. The performance of four processing methods derived under low signal-to-noise ratio (SNR) conditions are analyzed and compared. Then, the multipath localization problem is considered for the case where there is a lack of precise knowledge of the propagation environment. Multipath localization based on a combination of a matched field type processing technique and recent superresolution techniques in adaptive array processing is reported. Simulations for localization are performed based on data obtained from the generic sonar model (GSM).>

Multidimensional generalizations of various 1-D results on the robustness of Hurwitz, Schur, and positivity properties of polynomials and rational functions are considered. More specifically, the convexity property of the stable region in the coefficient space of multivariable polynomials is studied. Multidimensional generalizations of Kharitonov-type results are reviewed, and further extensions, including that of the 1-D edge theorem, are discussed. Interval positivity properties of multivariate rational functions are characterized in terms of ratios of a finite number of Kharitonov-type polynomials constructed from the extreme values of the intervals of perturbation.