A simple form of the linguist. It makes the following simplifying assumptions: 1) Zero or one word per grammar node 2) No fan-in allowed ever 3) No co

An abstract decoder which implements all functionality which is independent of the used decoding-paradigm (pull/push).

Abstract base class for windowed feature extractors like DeltasFeatureExtractor, ConcatFeatureExtractor or S3FeatureExtractor. The main purpose of thi

Base class for high-level speech recognizers.

An abstract analyzer that signals about presense of speech in last processing frame. This information is used in noise filtering components to estimat

Tracks and reports recognition accuracy

Represents the generic interface to the Acoustic Model for sphinx4

Provides a mechanism for scoring a set of HMM states

An active list is maintained as a sorted list <p> Note that all scores are represented in LogMath logbase

Creates new active lists.

An active list is maintained as a sorted list

This is a simple tool to align audio to text and dump a database for the training/evaluation. You need to provide a model, dictionary, audio and the

This class demonstrates how to align audio to existing transcription and receive word timestamps. <br> In order to initialize the aligner you need to

A simple example that shows how to transcribe a continuous audio file that has multiple utterances in it.

Manager for pruned hypothesis @author Joe Woelfel

Represents a set of alternatives in an SentenceHMMS

DOCUMENT ME! @author Holger Brandl

The fst's arc implementation. @author John Salatas

ArcSort operation. @author John Salatas

{@code FrontEnd} element that asserts the audio-stream to be continuous. This is often a mandatory property for frontend setups. The component operate

Represents a 16bit, SIGNED_PCM, little endian audio clip with a sample rate specified by AudioFormat.

Converts an AudioData into an InputStream.

Some small unit tests to check whether the AudioFileDataSource and the ConcatAudioFileDataSource are working properly.

An AudioFileDataSource generates a stream of audio data from a given audio file. All required information concerning the audio format are read directl

An interface which is describes the functionality which is required to handle new file signals fired by the audio-data sources. @author Holger Brandl

Provides an interface to view and play back various forms of an audio signal.

Plays an AudioData in a separate thread.

Records and displays the waveform and spectrogram of an audio signal. See <a href="doc-files/HowToRunAudioTool.html">How to Run AudioTool</a> for inf

Cepstrum is an auto-configurable DataProcessor which is used to compute a specific cepstrum (for a target acoustic model) given the spectrum. The Ceps

Normalizes a set of Tokens against the best scoring Token of a background model. @author Holger Brandl

A simple energy-based detector for upsampled audio. Could be used to detect bandwidth issues leading to the accuracy issues. The detector simply look

An abstract DataProcessor implementing elements common to all concrete DataProcessors, such as name, predecessor, and timer.

Applies automatic gain control (CMN)

Applies cepstral mean normalization (CMN), sometimes called channel mean normalization, to incoming cepstral data. Its goal is to reduce the distorti

Batch decoder recognizer which can be used inside Sphinxtrain to build lattices and generate decoding results.

Provides a set of utilities methods for manipulating batch files.

@author Peter Wolf

Copyright 1999-2002 Carnegie Mellon University. Portions Copyright 2002 Sun Microsystems, Inc. Portions Copyright 2002 Mitsubishi Electric Research La

Provides a standard interface to for a single decode in a batch of decodes

Provides a standard interface to a batching mechanism

Decodes a batch file containing a list of files to decode. The files can be either audio files or cepstral files, but defaults to audio files. The aud

Copyright 1999-2002 Carnegie Mellon University. Portions Copyright 2002 Sun Microsystems, Inc. Portions Copyright 2002 Mitsubishi Electric Research La

Applies cepstral variance normalization (CVN), so that each coefficient will have unit variance. You need to put this element after the means normaliz

Provides mechanisms for computing statistics given a set of states and input data.

Tracks and reports recognition accuracy based upon the highest scoring path in a Result.

Reads a binary NGram language model file ("DMP file") generated by the SphinxBase sphinx_lm_convert. <p> Note that all probabilities in the grammar ar

Class that provides utils to load NgramTrieModel from binary file created with sphinx_lm_convert. Routines should be called in certain order according

Language model that reads whole model into memory. Useful for loading language models from resources or external locations. @author Nikolay Shmyrev

Represents a branching node in a grammar

Used to accumulate data for updating of models.

This class builds an HMM from a transcript, at increasing levels of details.

Constructs a loop of all the context-independent phones. This loop is used in the static flat linguist for detecting out-of-grammar utterances. A 'pho

Some unit-tests for the ConfigurationManagerUtils.

Provides a simple object cache. <p>Object stored in cache must properly implement {@link Object#hashCode hashCode} and {@link Object#equals equals}.

Converts a set of log magnitude Spectrum data into a graphical representation.

Implements the tokenizer interface. Breaks an input sequence of characters into a set of tokens.

An LM that computes a probability of a word sequence by converting words to classes and asking the class-based probability from a delegate LM. @autho

A component that knows how to map words to classes and vice versa. @author Tanel Alumae

The only purpose of this class is to present all classes as words in the getWord() method. This dictionary must be injected to a class LM as a diction

Represents a probability of a word belonging to class. @author Tanel Alumae

Used for clustering gaussians. The clustering is performed by Euclidean distance criterion. The "k-means" clustering algorithm is used for clustering

A Color is used to tag SentenceHMM nodes

An interface implemented by command functions typically added to a command interpreter @see CommandInterpreter

This class is a command interpreter. It reads strings from an input stream, parses them into commands and executes them, results are sent back on the

A decision Node that compares two values.

Implements complex types and arythmetics

DOCUMENT ME! @author Holger Brandl

DOCUMENT ME! @author Holger Brandl

Compose operation. See: M. Mohri, "Weighted automata algorithms", Handbook of Weighted Automata. Springer, pp. 213-250, 2009. @author John Salatas

@author John Salatas

Compose Testing for Examples provided by M. Mohri, "Weighted Automata Algorithms", Handbook of Weighted Automata, Springer-Verlag, 2009, pp. 213-254.

Represents a composite senone. A composite senone consists of a set of all possible {@link Senone senones} for a given state. CompositeSenones are use

Concatenates a list of audio files as one continuous audio stream. <p> A {@link edu.cmu.sphinx.frontend.DataStartSignal DataStartSignal} will be place

This component concatenate the cepstrum from the sequence of frames according to the window size. It's not supposed to give high accuracy alone, but c

Concatenates a list raw headerless audio files as one continuous audio stream. A {@link edu.cmu.sphinx.frontend.DataStartSignal DataStartSignal} will

A SAX XML Handler implementation that builds up the map of raw property data objects

Shows the configuration currently in use. This monitor is typically added as a recognition monitor such that the configuration is shown immediately af

Defines the interface that must be implemented by any configurable component in Sphinx-4. The life cycle of a component is as follows: <ul><li> <b>Cl

An default (abstract) implementation of a configurable that implements a meaning {@code toString()} and keeps a references to the {@code Confurable}'s

Represents common configuration options. This configuration is used by high-level recognition classes. @see SpeechAligner @see LiveSpeechRecognizer

Describes all methods necessary to process change events of a <code>ConfigurationManager</code>. @author Holger Brandl @see edu.cmu.sphinx.util.props

Manages a set of <code>Configurable</code>s, their parameterization and the relationships between them. Configurations can be specified either by xml

Some unit tests, which ensure a proper implementation of configuration management. @author Holger Brandl