Continuous Speech Recognition Based Cognitive semi-supervised continuous learning View Homepage


Ontology type: schema:MonetaryGrant     


Grant Info

YEARS

2008-2010

FUNDING AMOUNT

280000 CNY

ABSTRACT

Considering the problem of efficiently using large quantity of non-labeled training data the continuous speech recognition system and improving the system’s self-adaptive learning ability, this project studied the system design methods and semi-supervised machine learning methods for the large vocabulary continuous speech recognition system construction, the learning learning models and methods through human cognition mechanism, the sequential data processing methods and machine learning methods, and the development of a prototype experimental system. The outputs of this project include: semi-supervised machine learning methods are proposed based on cognitive mechanism, through introducing grammatical rules, information compression methods and multi-stage BIC criteria etc., effectively realized the semi-supervised incremental learning approaches for LVASR system. Semi-supervised learning model and methods based on human cognition mechanism and learning mechanisms was established, the semi-supervised clustering algorithm, parallel learning methods and Co-Forest methods etc. are proposed based on Tri-training, data editing and MapReduce. In areas of sequential data processing, a mechanism was proposed based on human's music cognition mechanism, solved the problems of the melody detection, tonal analysis and classi More... »

URL

http://npd.nsfc.gov.cn/projectDetail.action?pid=60772076

Related SciGraph Publications

  • 2011-06. A new co-training-style random forest for computer aided diagnosis in JOURNAL OF INTELLIGENT INFORMATION SYSTEMS
  • JSON-LD is the canonical representation for SciGraph data.

    TIP: You can open this SciGraph record using an external JSON-LD service: JSON-LD Playground Google SDTT

    [
      {
        "@context": "https://springernature.github.io/scigraph/jsonld/sgcontext.json", 
        "about": [
          {
            "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/2217", 
            "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
            "type": "DefinedTerm"
          }, 
          {
            "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/2217", 
            "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
            "type": "DefinedTerm"
          }, 
          {
            "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/2208", 
            "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
            "type": "DefinedTerm"
          }
        ], 
        "amount": {
          "currency": "CNY", 
          "type": "MonetaryAmount", 
          "value": "280000"
        }, 
        "description": "Considering the problem of efficiently using large quantity of non-labeled training data the continuous speech recognition system and improving the system\u2019s self-adaptive learning ability, this project studied the system design methods and semi-supervised machine learning methods for the large vocabulary continuous speech recognition system construction, the learning learning models and methods through human cognition mechanism, the sequential data processing methods and machine learning methods, and the development of a prototype experimental system.\nThe outputs of this project include: semi-supervised machine learning methods are proposed based on cognitive mechanism, through introducing grammatical rules, information compression methods and multi-stage BIC criteria etc., effectively realized the semi-supervised incremental learning approaches for LVASR system. Semi-supervised learning model and methods based on human cognition mechanism and learning mechanisms was established, the semi-supervised clustering algorithm, parallel learning methods and Co-Forest methods etc. are proposed based on Tri-training, data editing and MapReduce. In areas of sequential data processing, a mechanism was proposed based on human's music cognition mechanism, solved the problems of the melody detection, tonal analysis and classi", 
        "endDate": "2010-12-31T00:00:00Z", 
        "funder": {
          "id": "https://www.grid.ac/institutes/grid.419696.5", 
          "type": "Organization"
        }, 
        "id": "sg:grant.4903809", 
        "identifier": [
          {
            "name": "dimensions_id", 
            "type": "PropertyValue", 
            "value": [
              "4903809"
            ]
          }, 
          {
            "name": "nsfc_id", 
            "type": "PropertyValue", 
            "value": [
              "60772076"
            ]
          }
        ], 
        "inLanguage": [
          "zh"
        ], 
        "keywords": [
          "machine", 
          "problem", 
          "LVASR system", 
          "melody detection", 
          "grammatical rules", 
          "continuous speech recognition system", 
          "experimental system", 
          "machine learning methods", 
          "clustering algorithm", 
          "METHODS", 
          "approach", 
          "development", 
          "output", 
          "semi", 
          "Tri-training", 
          "non-labeled training data", 
          "semi-supervised machine learning methods", 
          "continuous speech recognition system construction", 
          "continuous learning", 
          "learning learning models", 
          "human cognition mechanism", 
          "sequential data processing methods", 
          "MapReduce", 
          "sequential data processing", 
          "area", 
          "system\u2019s self-adaptive learning ability", 
          "cognitive mechanisms", 
          "human's music cognition mechanism", 
          "data editing", 
          "Co-Forest methods", 
          "parallel learning methods", 
          "classi", 
          "learning model", 
          "system design method", 
          "large quantities", 
          "prototype", 
          "information compression methods", 
          "Continuous Speech Recognition Based Cognitive semi", 
          "multi-stage BIC criteria", 
          "mechanism", 
          "large vocabulary", 
          "tonal analysis", 
          "project", 
          "learning mechanism"
        ], 
        "name": "Continuous Speech Recognition Based Cognitive semi-supervised continuous learning", 
        "recipient": [
          {
            "id": "https://www.grid.ac/institutes/grid.19373.3f", 
            "type": "Organization"
          }, 
          {
            "affiliation": {
              "id": "https://www.grid.ac/institutes/grid.19373.3f", 
              "name": "Harbin Institute of Technology", 
              "type": "Organization"
            }, 
            "familyName": "Li", 
            "givenName": "Hai Feng", 
            "id": "sg:person.0603055335.10", 
            "type": "Person"
          }, 
          {
            "member": "sg:person.0603055335.10", 
            "roleName": "PI", 
            "type": "Role"
          }
        ], 
        "sameAs": [
          "https://app.dimensions.ai/details/grant/grant.4903809"
        ], 
        "sdDataset": "grants", 
        "sdDatePublished": "2019-03-07T12:43", 
        "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
        "sdPublisher": {
          "name": "Springer Nature - SN SciGraph project", 
          "type": "Organization"
        }, 
        "sdSource": "s3://com.uberresearch.data.processor/core_data/20181219_192338/projects/base/nsfc_projects_5.xml.gz", 
        "startDate": "2008-01-01T00:00:00Z", 
        "type": "MonetaryGrant", 
        "url": "http://npd.nsfc.gov.cn/projectDetail.action?pid=60772076"
      }
    ]
     

    Download the RDF metadata as:  json-ld nt turtle xml License info

    HOW TO GET THIS DATA PROGRAMMATICALLY:

    JSON-LD is a popular format for linked data which is fully compatible with JSON.

    curl -H 'Accept: application/ld+json' 'https://scigraph.springernature.com/grant.4903809'

    N-Triples is a line-based linked data format ideal for batch operations.

    curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/grant.4903809'

    Turtle is a human-readable linked data format.

    curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/grant.4903809'

    RDF/XML is a standard XML format for linked data.

    curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/grant.4903809'


     

    This table displays all metadata directly associated to this object as RDF triples.

    91 TRIPLES      19 PREDICATES      67 URIs      58 LITERALS      5 BLANK NODES

    Subject Predicate Object
    1 sg:grant.4903809 schema:about anzsrc-for:2208
    2 anzsrc-for:2217
    3 schema:amount Ndb75376f226f48618a341a650a97706a
    4 schema:description Considering the problem of efficiently using large quantity of non-labeled training data the continuous speech recognition system and improving the system’s self-adaptive learning ability, this project studied the system design methods and semi-supervised machine learning methods for the large vocabulary continuous speech recognition system construction, the learning learning models and methods through human cognition mechanism, the sequential data processing methods and machine learning methods, and the development of a prototype experimental system. The outputs of this project include: semi-supervised machine learning methods are proposed based on cognitive mechanism, through introducing grammatical rules, information compression methods and multi-stage BIC criteria etc., effectively realized the semi-supervised incremental learning approaches for LVASR system. Semi-supervised learning model and methods based on human cognition mechanism and learning mechanisms was established, the semi-supervised clustering algorithm, parallel learning methods and Co-Forest methods etc. are proposed based on Tri-training, data editing and MapReduce. In areas of sequential data processing, a mechanism was proposed based on human's music cognition mechanism, solved the problems of the melody detection, tonal analysis and classi
    5 schema:endDate 2010-12-31T00:00:00Z
    6 schema:funder https://www.grid.ac/institutes/grid.419696.5
    7 schema:identifier N84cf996817f649a497367e6e3b29a56d
    8 Ne9ab81535bd1429791e8733bed79dedc
    9 schema:inLanguage zh
    10 schema:keywords Co-Forest methods
    11 Continuous Speech Recognition Based Cognitive semi
    12 LVASR system
    13 METHODS
    14 MapReduce
    15 Tri-training
    16 approach
    17 area
    18 classi
    19 clustering algorithm
    20 cognitive mechanisms
    21 continuous learning
    22 continuous speech recognition system
    23 continuous speech recognition system construction
    24 data editing
    25 development
    26 experimental system
    27 grammatical rules
    28 human cognition mechanism
    29 human's music cognition mechanism
    30 information compression methods
    31 large quantities
    32 large vocabulary
    33 learning learning models
    34 learning mechanism
    35 learning model
    36 machine
    37 machine learning methods
    38 mechanism
    39 melody detection
    40 multi-stage BIC criteria
    41 non-labeled training data
    42 output
    43 parallel learning methods
    44 problem
    45 project
    46 prototype
    47 semi
    48 semi-supervised machine learning methods
    49 sequential data processing
    50 sequential data processing methods
    51 system design method
    52 system’s self-adaptive learning ability
    53 tonal analysis
    54 schema:name Continuous Speech Recognition Based Cognitive semi-supervised continuous learning
    55 schema:recipient Ncf2f0eaba5894183814b9b0ab8667f82
    56 sg:person.0603055335.10
    57 https://www.grid.ac/institutes/grid.19373.3f
    58 schema:sameAs https://app.dimensions.ai/details/grant/grant.4903809
    59 schema:sdDatePublished 2019-03-07T12:43
    60 schema:sdLicense https://scigraph.springernature.com/explorer/license/
    61 schema:sdPublisher Neac805f5d6034037b631511b3d6afef3
    62 schema:startDate 2008-01-01T00:00:00Z
    63 schema:url http://npd.nsfc.gov.cn/projectDetail.action?pid=60772076
    64 sgo:license sg:explorer/license/
    65 sgo:sdDataset grants
    66 rdf:type schema:MonetaryGrant
    67 N84cf996817f649a497367e6e3b29a56d schema:name nsfc_id
    68 schema:value 60772076
    69 rdf:type schema:PropertyValue
    70 Ncf2f0eaba5894183814b9b0ab8667f82 schema:member sg:person.0603055335.10
    71 schema:roleName PI
    72 rdf:type schema:Role
    73 Ndb75376f226f48618a341a650a97706a schema:currency CNY
    74 schema:value 280000
    75 rdf:type schema:MonetaryAmount
    76 Ne9ab81535bd1429791e8733bed79dedc schema:name dimensions_id
    77 schema:value 4903809
    78 rdf:type schema:PropertyValue
    79 Neac805f5d6034037b631511b3d6afef3 schema:name Springer Nature - SN SciGraph project
    80 rdf:type schema:Organization
    81 anzsrc-for:2208 schema:inDefinedTermSet anzsrc-for:
    82 rdf:type schema:DefinedTerm
    83 anzsrc-for:2217 schema:inDefinedTermSet anzsrc-for:
    84 rdf:type schema:DefinedTerm
    85 sg:person.0603055335.10 schema:affiliation https://www.grid.ac/institutes/grid.19373.3f
    86 schema:familyName Li
    87 schema:givenName Hai Feng
    88 rdf:type schema:Person
    89 https://www.grid.ac/institutes/grid.19373.3f schema:name Harbin Institute of Technology
    90 rdf:type schema:Organization
    91 https://www.grid.ac/institutes/grid.419696.5 schema:Organization
     




    Preview window. Press ESC to close (or click here)


    ...