The Experts below are selected from a list of 25872 Experts worldwide ranked by ideXlab platform

Sanjoy K Mitter - One of the best experts on this subject based on the ideXlab platform.

  • a Linguistic Feature representation of the speech waveform
    International Conference on Acoustics Speech and Signal Processing, 1993
    Co-Authors: E Eide, J R Rohlicek, H Gish, Sanjoy K Mitter
    Abstract:

    Linguistic theory views a phoneme as a shorthand notation for a bundle of binary Features related to the operation of the speaker's articulators. A representation of the speech waveform in terms of these underlying distinctive Features is described here. The estimation of the probability of each of 14 Linguistic Features being encoded locally in the waveform is performed on a frame-by-frame basis. In going from the abstract to the physical level, it is recognized that the Features are encoded in the waveform hierarchically and that time-varying manifestations of a Feature within a phonemic segment are possible. These issues are addressed simultaneously through a two-stage procedure. In the first pass, the time portion and broad class of sound being represented by each frame are estimated. On the second pass, for each distinctive Linguistic Feature, models built explicitly for the estimated broad class portion are evaluated to arrive at the probability that each frame is part of a realization of a phoneme in which the Feature is present. The distinctive Feature representation is applied to the tasks of phoneme recognition and secondary classification in keyword spotting. >

  • ICASSP (2) - A Linguistic Feature representation of the speech waveform
    IEEE International Conference on Acoustics Speech and Signal Processing, 1993
    Co-Authors: E Eide, J R Rohlicek, H Gish, Sanjoy K Mitter
    Abstract:

    Linguistic theory views a phoneme as a shorthand notation for a bundle of binary Features related to the operation of the speaker's articulators. A representation of the speech waveform in terms of these underlying distinctive Features is described here. The estimation of the probability of each of 14 Linguistic Features being encoded locally in the waveform is performed on a frame-by-frame basis. In going from the abstract to the physical level, it is recognized that the Features are encoded in the waveform hierarchically and that time-varying manifestations of a Feature within a phonemic segment are possible. These issues are addressed simultaneously through a two-stage procedure. In the first pass, the time portion and broad class of sound being represented by each frame are estimated. On the second pass, for each distinctive Linguistic Feature, models built explicitly for the estimated broad class portion are evaluated to arrive at the probability that each frame is part of a realization of a phoneme in which the Feature is present. The distinctive Feature representation is applied to the tasks of phoneme recognition and secondary classification in keyword spotting. >

Darijus Strasunskas - One of the best experts on this subject based on the ideXlab platform.

  • semantic Linguistic Feature vectors for search unsupervised construction and experimental validation
    Asian Semantic Web Conference, 2009
    Co-Authors: Stein L Tomassen, Darijus Strasunskas
    Abstract:

    In this paper, we elaborate on an approach to construction of semantic-Linguistic Feature vectors (FV) that are used in search. These FVs are built based on domain semantics encoded in an ontology and enhanced by a relevant terminology from Web documents. The value of this approach is twofold. First, it captures relevant semantics from an ontology, and second, it accounts for statistically significant collocations of other terms and phrases in relation to the ontology entities. The contribution of this paper is the FV construction process and its evaluation. Recommendations and lessons learnt are laid down.

  • ASWC - Semantic-Linguistic Feature Vectors for Search: Unsupervised Construction and Experimental Validation
    The Semantic Web, 2009
    Co-Authors: Stein L Tomassen, Darijus Strasunskas
    Abstract:

    In this paper, we elaborate on an approach to construction of semantic-Linguistic Feature vectors (FV) that are used in search. These FVs are built based on domain semantics encoded in an ontology and enhanced by a relevant terminology from Web documents. The value of this approach is twofold. First, it captures relevant semantics from an ontology, and second, it accounts for statistically significant collocations of other terms and phrases in relation to the ontology entities. The contribution of this paper is the FV construction process and its evaluation. Recommendations and lessons learnt are laid down.

  • Construction of Ontology Based Semantic-Linguistic Feature Vectors for Searching: The Process and Effect
    2009 IEEE WIC ACM International Joint Conference on Web Intelligence and Intelligent Agent Technology, 2009
    Co-Authors: Stein L Tomassen, Darijus Strasunskas
    Abstract:

    Search is among the most frequent activities on the Web. However, the search activity still requires extra efforts in order to get satisfactory results. One of the reasons is heterogeneous information resources and exponential growth of information. The problem of heterogeneity arises as a result of discipline specific language used even in domain specific documents. This particular problem we tackle in this paper. We propose an approach to construct semantic-Linguistic Feature vectors (FV). The FVs are built based on domain semantics encoded in an ontology and enhanced by a relevant terminology from documents on the Web. Semantic information from the ontologies is also used to expand the user queries and the FVs are used to filter and rank the retrieved documents. The strength of this approach is twofold. First, it is grounded on relevant semantics from an ontology, and second, it accounts for statistically significant collocations of other terms and phrases in relation to the ontology entities. In this paper, we explain how these FVs are constructed and what effect they have on search performance.

Richard Bowden - One of the best experts on this subject based on the ideXlab platform.

  • a Linguistic Feature vector for the visual interpretation of sign language
    European Conference on Computer Vision, 2004
    Co-Authors: Richard Bowden, David Windridge, Timor Kadir, Andrew Zisserman, Michael Brady
    Abstract:

    This paper presents a novel approach to sign language recognition that provides extremely high classification rates on minimal training data. Key to this approach is a 2 stage classification procedure where an initial classification stage extracts a high level description of hand shape and motion. This high level description is based upon sign Linguistics and describes actions at a conceptual level easily understood by humans. Moreover, such a description broadly generalises temporal activities naturally overcoming variability of people and environments. A second stage of classification is then used to model the temporal transitions of individual signs using a classifier bank of Markov chains combined with Independent Component Analysis. We demonstrate classification rates as high as 97.67% for a lexicon of 43 words using only single instance training outperforming previous approaches where thousands of training examples are required.

  • ECCV (1) - A Linguistic Feature Vector for the Visual Interpretation of Sign Language
    Lecture Notes in Computer Science, 2004
    Co-Authors: Richard Bowden, David Windridge, Timor Kadir, Andrew Zisserman, Michael Brady
    Abstract:

    This paper presents a novel approach to sign language recognition that provides extremely high classification rates on minimal training data. Key to this approach is a 2 stage classification procedure where an initial classification stage extracts a high level description of hand shape and motion. This high level description is based upon sign Linguistics and describes actions at a conceptual level easily understood by humans. Moreover, such a description broadly generalises temporal activities naturally overcoming variability of people and environments. A second stage of classification is then used to model the temporal transitions of individual signs using a classifier bank of Markov chains combined with Independent Component Analysis. We demonstrate classification rates as high as 97.67% for a lexicon of 43 words using only single instance training outperforming previous approaches where thousands of training examples are required.

Michael Brady - One of the best experts on this subject based on the ideXlab platform.

  • a Linguistic Feature vector for the visual interpretation of sign language
    European Conference on Computer Vision, 2004
    Co-Authors: Richard Bowden, David Windridge, Timor Kadir, Andrew Zisserman, Michael Brady
    Abstract:

    This paper presents a novel approach to sign language recognition that provides extremely high classification rates on minimal training data. Key to this approach is a 2 stage classification procedure where an initial classification stage extracts a high level description of hand shape and motion. This high level description is based upon sign Linguistics and describes actions at a conceptual level easily understood by humans. Moreover, such a description broadly generalises temporal activities naturally overcoming variability of people and environments. A second stage of classification is then used to model the temporal transitions of individual signs using a classifier bank of Markov chains combined with Independent Component Analysis. We demonstrate classification rates as high as 97.67% for a lexicon of 43 words using only single instance training outperforming previous approaches where thousands of training examples are required.

  • ECCV (1) - A Linguistic Feature Vector for the Visual Interpretation of Sign Language
    Lecture Notes in Computer Science, 2004
    Co-Authors: Richard Bowden, David Windridge, Timor Kadir, Andrew Zisserman, Michael Brady
    Abstract:

    This paper presents a novel approach to sign language recognition that provides extremely high classification rates on minimal training data. Key to this approach is a 2 stage classification procedure where an initial classification stage extracts a high level description of hand shape and motion. This high level description is based upon sign Linguistics and describes actions at a conceptual level easily understood by humans. Moreover, such a description broadly generalises temporal activities naturally overcoming variability of people and environments. A second stage of classification is then used to model the temporal transitions of individual signs using a classifier bank of Markov chains combined with Independent Component Analysis. We demonstrate classification rates as high as 97.67% for a lexicon of 43 words using only single instance training outperforming previous approaches where thousands of training examples are required.

E Eide - One of the best experts on this subject based on the ideXlab platform.

  • a Linguistic Feature representation of the speech waveform
    International Conference on Acoustics Speech and Signal Processing, 1993
    Co-Authors: E Eide, J R Rohlicek, H Gish, Sanjoy K Mitter
    Abstract:

    Linguistic theory views a phoneme as a shorthand notation for a bundle of binary Features related to the operation of the speaker's articulators. A representation of the speech waveform in terms of these underlying distinctive Features is described here. The estimation of the probability of each of 14 Linguistic Features being encoded locally in the waveform is performed on a frame-by-frame basis. In going from the abstract to the physical level, it is recognized that the Features are encoded in the waveform hierarchically and that time-varying manifestations of a Feature within a phonemic segment are possible. These issues are addressed simultaneously through a two-stage procedure. In the first pass, the time portion and broad class of sound being represented by each frame are estimated. On the second pass, for each distinctive Linguistic Feature, models built explicitly for the estimated broad class portion are evaluated to arrive at the probability that each frame is part of a realization of a phoneme in which the Feature is present. The distinctive Feature representation is applied to the tasks of phoneme recognition and secondary classification in keyword spotting. >

  • ICASSP (2) - A Linguistic Feature representation of the speech waveform
    IEEE International Conference on Acoustics Speech and Signal Processing, 1993
    Co-Authors: E Eide, J R Rohlicek, H Gish, Sanjoy K Mitter
    Abstract:

    Linguistic theory views a phoneme as a shorthand notation for a bundle of binary Features related to the operation of the speaker's articulators. A representation of the speech waveform in terms of these underlying distinctive Features is described here. The estimation of the probability of each of 14 Linguistic Features being encoded locally in the waveform is performed on a frame-by-frame basis. In going from the abstract to the physical level, it is recognized that the Features are encoded in the waveform hierarchically and that time-varying manifestations of a Feature within a phonemic segment are possible. These issues are addressed simultaneously through a two-stage procedure. In the first pass, the time portion and broad class of sound being represented by each frame are estimated. On the second pass, for each distinctive Linguistic Feature, models built explicitly for the estimated broad class portion are evaluated to arrive at the probability that each frame is part of a realization of a phoneme in which the Feature is present. The distinctive Feature representation is applied to the tasks of phoneme recognition and secondary classification in keyword spotting. >