The Experts below are selected from a list of 93 Experts worldwide ranked by ideXlab platform

Takuya Higuchi - One of the best experts on this subject based on the ideXlab platform.

  • the ntt chime 3 system advances in speech enhancement and recognition for mobile multi microphone devices
    IEEE Automatic Speech Recognition and Understanding Workshop, 2015
    Co-Authors: Takuya Yoshioka, Nobutaka Ito, Marc Delcroix, Atsunori Ogawa, Keisuke Kinoshita, Masakiyo Fujimoto, Wojciech J Fabian, Miquel Espi, Takuya Higuchi, Shoko Araki
    Abstract:

    CHiME-3 is a research community challenge organised in 2015 to evaluate speech recognition systems for mobile multi-microphone devices used in noisy daily environments. This paper describes NTT's CHiME-3 system, which integrates advanced speech enhancement and recognition techniques. Newly developed techniques include the use of spectral masks for acoustic beam-steering vector estimation and acoustic modelling with deep convolutional neural networks based on the "network in network" concept. In addition to these improvements, our system has several key differences from the official baseline system. The differences include multi-microphone training, dereverberation, and cross adaptation of neural networks with different architectures. The impacts that these techniques have on recognition performance are investigated. By combining these advanced techniques, our system achieves a 3.45% Development Error rate and a 5.83% evaluation Error rate. Three simpler systems are also developed to perform evaluations with constrained set-ups.

  • ASRU - The NTT CHiME-3 system: Advances in speech enhancement and recognition for mobile multi-microphone devices
    2015 IEEE Workshop on Automatic Speech Recognition and Understanding (ASRU), 2015
    Co-Authors: Takuya Yoshioka, Nobutaka Ito, Marc Delcroix, Atsunori Ogawa, Keisuke Kinoshita, Masakiyo Fujimoto, Wojciech J Fabian, Miquel Espi, Takuya Higuchi
    Abstract:

    CHiME-3 is a research community challenge organised in 2015 to evaluate speech recognition systems for mobile multi-microphone devices used in noisy daily environments. This paper describes NTT's CHiME-3 system, which integrates advanced speech enhancement and recognition techniques. Newly developed techniques include the use of spectral masks for acoustic beam-steering vector estimation and acoustic modelling with deep convolutional neural networks based on the "network in network" concept. In addition to these improvements, our system has several key differences from the official baseline system. The differences include multi-microphone training, dereverberation, and cross adaptation of neural networks with different architectures. The impacts that these techniques have on recognition performance are investigated. By combining these advanced techniques, our system achieves a 3.45% Development Error rate and a 5.83% evaluation Error rate. Three simpler systems are also developed to perform evaluations with constrained set-ups.

Laura Nicholson - One of the best experts on this subject based on the ideXlab platform.

  • Size Sequencing as a Window on Executive Control in Children with Autism and Asperger’s Syndrome
    Journal of Autism and Developmental Disorders, 2007
    Co-Authors: Margaret Mcgonigle-chalmers, Kimberly Bodner, Alicia Fox-pitt, Laura Nicholson
    Abstract:

    A study is reported in which size sequencing on a touch screen is used as a measure of executive control in 20 high-functioning children with Autistic Spectrum Disorders (ASD). The data show a significant and age-independent effect of the length of sequence that can be executed without Errors by these children, in comparison with a chronologically age-matched group of children with normal Development. Error data and reaction times are analysed and are interpreted as revealing a constraint on the prospective component of working memory in children on the autistic spectrum even when there is no change in goal or perceptual set. It is concluded that the size sequencing paradigm is an effective measure of executive difficulties associated with autism.

  • Size Sequencing as a Window on Executive Control in Children with Autism and Asperger's Syndrome
    Journal of autism and developmental disorders, 2007
    Co-Authors: Margaret Mcgonigle-chalmers, Alicia Fox-pitt, Kimberly E. Bodner, Laura Nicholson
    Abstract:

    A study is reported in which size sequencing on a touch screen is used as a measure of executive control in 20 high-functioning children with Autistic Spectrum Disorders (ASD). The data show a significant and age-independent effect of the length of sequence that can be executed without Errors by these children, in comparison with a chronologically age-matched group of children with normal Development. Error data and reaction times are analysed and are interpreted as revealing a constraint on the prospective component of working memory in children on the autistic spectrum even when there is no change in goal or perceptual set. It is concluded that the size sequencing paradigm is an effective measure of executive difficulties associated with autism.

Takuya Yoshioka - One of the best experts on this subject based on the ideXlab platform.

  • the ntt chime 3 system advances in speech enhancement and recognition for mobile multi microphone devices
    IEEE Automatic Speech Recognition and Understanding Workshop, 2015
    Co-Authors: Takuya Yoshioka, Nobutaka Ito, Marc Delcroix, Atsunori Ogawa, Keisuke Kinoshita, Masakiyo Fujimoto, Wojciech J Fabian, Miquel Espi, Takuya Higuchi, Shoko Araki
    Abstract:

    CHiME-3 is a research community challenge organised in 2015 to evaluate speech recognition systems for mobile multi-microphone devices used in noisy daily environments. This paper describes NTT's CHiME-3 system, which integrates advanced speech enhancement and recognition techniques. Newly developed techniques include the use of spectral masks for acoustic beam-steering vector estimation and acoustic modelling with deep convolutional neural networks based on the "network in network" concept. In addition to these improvements, our system has several key differences from the official baseline system. The differences include multi-microphone training, dereverberation, and cross adaptation of neural networks with different architectures. The impacts that these techniques have on recognition performance are investigated. By combining these advanced techniques, our system achieves a 3.45% Development Error rate and a 5.83% evaluation Error rate. Three simpler systems are also developed to perform evaluations with constrained set-ups.

  • ASRU - The NTT CHiME-3 system: Advances in speech enhancement and recognition for mobile multi-microphone devices
    2015 IEEE Workshop on Automatic Speech Recognition and Understanding (ASRU), 2015
    Co-Authors: Takuya Yoshioka, Nobutaka Ito, Marc Delcroix, Atsunori Ogawa, Keisuke Kinoshita, Masakiyo Fujimoto, Wojciech J Fabian, Miquel Espi, Takuya Higuchi
    Abstract:

    CHiME-3 is a research community challenge organised in 2015 to evaluate speech recognition systems for mobile multi-microphone devices used in noisy daily environments. This paper describes NTT's CHiME-3 system, which integrates advanced speech enhancement and recognition techniques. Newly developed techniques include the use of spectral masks for acoustic beam-steering vector estimation and acoustic modelling with deep convolutional neural networks based on the "network in network" concept. In addition to these improvements, our system has several key differences from the official baseline system. The differences include multi-microphone training, dereverberation, and cross adaptation of neural networks with different architectures. The impacts that these techniques have on recognition performance are investigated. By combining these advanced techniques, our system achieves a 3.45% Development Error rate and a 5.83% evaluation Error rate. Three simpler systems are also developed to perform evaluations with constrained set-ups.

Wojciech J Fabian - One of the best experts on this subject based on the ideXlab platform.

  • the ntt chime 3 system advances in speech enhancement and recognition for mobile multi microphone devices
    IEEE Automatic Speech Recognition and Understanding Workshop, 2015
    Co-Authors: Takuya Yoshioka, Nobutaka Ito, Marc Delcroix, Atsunori Ogawa, Keisuke Kinoshita, Masakiyo Fujimoto, Wojciech J Fabian, Miquel Espi, Takuya Higuchi, Shoko Araki
    Abstract:

    CHiME-3 is a research community challenge organised in 2015 to evaluate speech recognition systems for mobile multi-microphone devices used in noisy daily environments. This paper describes NTT's CHiME-3 system, which integrates advanced speech enhancement and recognition techniques. Newly developed techniques include the use of spectral masks for acoustic beam-steering vector estimation and acoustic modelling with deep convolutional neural networks based on the "network in network" concept. In addition to these improvements, our system has several key differences from the official baseline system. The differences include multi-microphone training, dereverberation, and cross adaptation of neural networks with different architectures. The impacts that these techniques have on recognition performance are investigated. By combining these advanced techniques, our system achieves a 3.45% Development Error rate and a 5.83% evaluation Error rate. Three simpler systems are also developed to perform evaluations with constrained set-ups.

  • ASRU - The NTT CHiME-3 system: Advances in speech enhancement and recognition for mobile multi-microphone devices
    2015 IEEE Workshop on Automatic Speech Recognition and Understanding (ASRU), 2015
    Co-Authors: Takuya Yoshioka, Nobutaka Ito, Marc Delcroix, Atsunori Ogawa, Keisuke Kinoshita, Masakiyo Fujimoto, Wojciech J Fabian, Miquel Espi, Takuya Higuchi
    Abstract:

    CHiME-3 is a research community challenge organised in 2015 to evaluate speech recognition systems for mobile multi-microphone devices used in noisy daily environments. This paper describes NTT's CHiME-3 system, which integrates advanced speech enhancement and recognition techniques. Newly developed techniques include the use of spectral masks for acoustic beam-steering vector estimation and acoustic modelling with deep convolutional neural networks based on the "network in network" concept. In addition to these improvements, our system has several key differences from the official baseline system. The differences include multi-microphone training, dereverberation, and cross adaptation of neural networks with different architectures. The impacts that these techniques have on recognition performance are investigated. By combining these advanced techniques, our system achieves a 3.45% Development Error rate and a 5.83% evaluation Error rate. Three simpler systems are also developed to perform evaluations with constrained set-ups.

Nobutaka Ito - One of the best experts on this subject based on the ideXlab platform.

  • the ntt chime 3 system advances in speech enhancement and recognition for mobile multi microphone devices
    IEEE Automatic Speech Recognition and Understanding Workshop, 2015
    Co-Authors: Takuya Yoshioka, Nobutaka Ito, Marc Delcroix, Atsunori Ogawa, Keisuke Kinoshita, Masakiyo Fujimoto, Wojciech J Fabian, Miquel Espi, Takuya Higuchi, Shoko Araki
    Abstract:

    CHiME-3 is a research community challenge organised in 2015 to evaluate speech recognition systems for mobile multi-microphone devices used in noisy daily environments. This paper describes NTT's CHiME-3 system, which integrates advanced speech enhancement and recognition techniques. Newly developed techniques include the use of spectral masks for acoustic beam-steering vector estimation and acoustic modelling with deep convolutional neural networks based on the "network in network" concept. In addition to these improvements, our system has several key differences from the official baseline system. The differences include multi-microphone training, dereverberation, and cross adaptation of neural networks with different architectures. The impacts that these techniques have on recognition performance are investigated. By combining these advanced techniques, our system achieves a 3.45% Development Error rate and a 5.83% evaluation Error rate. Three simpler systems are also developed to perform evaluations with constrained set-ups.

  • ASRU - The NTT CHiME-3 system: Advances in speech enhancement and recognition for mobile multi-microphone devices
    2015 IEEE Workshop on Automatic Speech Recognition and Understanding (ASRU), 2015
    Co-Authors: Takuya Yoshioka, Nobutaka Ito, Marc Delcroix, Atsunori Ogawa, Keisuke Kinoshita, Masakiyo Fujimoto, Wojciech J Fabian, Miquel Espi, Takuya Higuchi
    Abstract:

    CHiME-3 is a research community challenge organised in 2015 to evaluate speech recognition systems for mobile multi-microphone devices used in noisy daily environments. This paper describes NTT's CHiME-3 system, which integrates advanced speech enhancement and recognition techniques. Newly developed techniques include the use of spectral masks for acoustic beam-steering vector estimation and acoustic modelling with deep convolutional neural networks based on the "network in network" concept. In addition to these improvements, our system has several key differences from the official baseline system. The differences include multi-microphone training, dereverberation, and cross adaptation of neural networks with different architectures. The impacts that these techniques have on recognition performance are investigated. By combining these advanced techniques, our system achieves a 3.45% Development Error rate and a 5.83% evaluation Error rate. Three simpler systems are also developed to perform evaluations with constrained set-ups.