Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Machine learning

Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Pattern Recognition and Machine Learning

Data Mining Practical Machine Learning Tools and Techniques

http://www.csun.edu/~andrzej/COMP529-S05/papers/Mesh%20Networks%20Survey.pdf

Wireless mesh networks: a survey

“Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks”の学習報告

In most real-world situations, a single microphone is insufficient for the characterization of an entire auditory scene. This often occurs in places such as office environments which consist of several interconnected spaces that are at least partially acoustically isolated from one another. To this end, we extend our previous work on segmentation of natural sounds to perform scene characterization using a sparse array of microphones, strategically placed to ensure that all parts of the environment are within range of at least one microphone. By accounting for which microphones are active for a given sound event, we perform a multi-channel segmentation that captures sound events occurring in any part of the space. The segmentation is inferred from a custom dynamic Bayesian network (DBN) that models how event boundaries influence changes in audio features. Example recordings illustrate the utility of our approach in a noisy office environment.

Multi-channel audio segmentation for continuous observation and archival of large spaces

Sinusoidal models are widely used in parametric speech and audio coding schemes. A common requirement in these applications is to select only a subset of components that provide the greatest perceptual benefit particularly at low bitrates. Usually, perceptual sinusoidal component selection algorithms make use of greedy algorithms that are computationally expensive. In this paper, we present a new algorithm that selects sinusoidal components based on the partial loudness model proposed by Moore & Glasberg. We compare the performance of the proposed algorithm in terms of perceptual benefit and computational complexity to other existing sinusoidal selection algorithms.

Sinusoidal component selection based on partial loudness criteria

In gaming and other multimedia applications, there is an increasing need to simulate and transform the sounds of physical interactions with real-world objects. When these sounds are based upon source recordings, this motivates the need for flexible, yet ecologically valid resyntheses. With modal analysis as a front end, banded waveguide resynthesis enables flexible modifications to dynamic spectral content and varied forms of interaction. Unfortunately, traditional banded waveguide models using biquad filters can have problems with undesired transient behavior. To this end, we introduce a new perfect reconstruction filterbank that allows for each filter to be individually tuned to a desired mode frequency and precisely cancel the other selected frequencies. Comparisons with traditional biquad models amply demonstrate the improved transient response and overall fidelity of our model. In addition, examples of resynthesis, transformation, and alternative interaction with real-world recordings are given.

An individually tunable perfect reconstruction filterbank for banded waveguide synthesis

The application of perceptual models in speech and audio coding began receiving attention during the late nineteen seventies. Methods that exploit the masking properties of the human ear in speech coding standards, even today, are largely based on relatively old concepts introduced by Schroeder and Atal in 1979. This dissertation studies a series of problems encountered in the application of new perceptual models in prediction-based speech and audio coding algorithms. The dissertation also explores different ways of integrating advanced human auditory models into fixed and variable bit rate standardized vocoders. 
Specific problems addressed in this dissertation include: (1) the significance of auditory excitation patterns in speech analysis and synthesis, (2) the performance of perceptual loudness metric in variable bit rate speech coding, and (3) adaptive pole estimation algorithms for linear prediction (LP) analysis in cascade form. The investigation of the above problems resulted in the development of two new algorithms for use in wideband speech coding and in variable bit rate speech coders. 
The first one is called the perceptually-motivated all-pole (PMAP) modeling algorithm. The PMAP algorithm is based on an approach for estimating the perceptually-relevant pole locations. The estimated perceptual poles are used to construct an all-pole filter for use in speech analysis. The proposed PMAP approach is compared against some of the existing perceptually-based linear prediction methods, i.e., the perceptual LP and the Warped LP. The PMAP modeling (1) provides a way to integrate psychoacoustic principles into LP by using auditory excitation pattern (AEP) matching, (2) enables estimation and perceptual ranking of the speech formants, and (3) provides an LP prediction residual with lower perceptual loudness. The computational profiling of the PMAP algorithm highlighted the modules that are computationally complex. In particular, the AEP-matching search contributed to the majority of the computational complexity. A fast-PMAP modeling that employs a block-form of AEP-matching was developed. By making use of the properties of the parametric spreading function and its energy-preserving smearing operation, the AEPs are recursively estimated. This recursive estimation of excitation patterns resulted in significant (over 50%) computational reduction. Experiments that compare the performance of the fast-PMAP algorithm relative to the original PMAP algorithm are included. 
The second algorithm is called the perceptual-loudness (PL) based rate determination. Unlike the existing rate selection strategies that are based on a voice activity detector and energy thresholds, the proposed method employs a perceptual loudness measure. The enhanced variable rate codec is used as the test-bed for evaluating the performance of the PL-based rate selection strategy. Experimental results demonstrate that the proposed PL-based rate determination compares well against other energy-based rate selection techniques in terms of average bit rate and speech quality. A fast PL-based rate selection algorithm that employs an LP analysis-driven pre-filtering followed by partial loudness estimation is proposed.

Andreas Spanias

Papers

Multi-channel audio segmentation for continuous observation and archival of large spaces

Sinusoidal component selection based on partial loudness criteria

An individually tunable perfect reconstruction filterbank for banded waveguide synthesis

Embedding perceptual linear prediction models in speech and audio coding

Audio Coding Standards and Algorithms