Learn More
Distant-microphone automatic speech recognition (ASR) remains a challenging goal in everyday environments involving multiple background sources and reverberation. This paper is intended to be a reference on the 2nd 'CHiME' Challenge, an initiative designed to analyze and evaluate the performance of ASR systems in a real-world domestic environment. Two(More)
Highly pathogenic avian H5N1 influenza A viruses occasionally infect humans, but currently do not transmit efficiently among humans. The viral haemagglutinin (HA) protein is a known host-range determinant as it mediates virus binding to host-specific cellular receptors. Here we assess the molecular changes in HA that would allow a virus possessing subtype(More)
We propose a new topic model for tracking time-varying consumer purchase behavior, in which consumer interests and item trends change over time. The proposed model can adaptively track changes in interests and trends based on current purchase logs and previously estimated interests and trends. The online nature of the proposed method means we do not need to(More)
The CHiME challenge series aims to advance far field speech recognition technology by promoting research at the interface of signal processing and automatic speech recognition. This paper presents the design and outcomes of the 3rd CHiME Challenge, which targets the performance of automatic speech recognition in a real-world, commercially-motivated(More)
This paper describes our joint efforts to provide robust automatic speech recognition (ASR) for reverberated environments, such as in hands-free human-machine interaction. We investigate blind feature space de-reverberation and deep recurrent de-noising auto-encoders (DAE) in an early fusion scheme. Results on the 2014 REVERB Challenge development set(More)
In this paper we propose Variational Bayesian Estimation and Clustering for speech recognition (VBEC), which is based on the Variational Bayesian (VB) approach. VBEC is a total Bayesian framework: all speech recognition procedures (acoustic modeling and speech classification) are based on VB posterior distribution, unlike the Maximum Likelihood (ML)(More)
Highly pathogenic avian H5N1 influenza A viruses have spread throughout Asia, Europe, and Africa, raising serious worldwide concern about their pandemic potential. Although more than 250 people have been infected with these viruses, with a consequent high rate of mortality, the molecular mechanisms responsible for the efficient transmission of H5N1 viruses(More)
Two amino acids (lysine at position 627 or asparagine at position 701) in the polymerase subunit PB2 protein are considered critical for the adaptation of avian influenza A viruses to mammals. However, the recently emerged pandemic H1N1 viruses lack these amino acids. Here, we report that a basic amino acid at position 591 of PB2 can compensate for the lack(More)
This paper proposes an activity recognition method that models an end user's activities without using any la-beled/unlabeled acceleration sensor data obtained from the user. Our method employs information about the end user's physical characteristics such as height and gender to find other users whose sensor data prepared in advance may be similar to those(More)
  • Marc Delcroix, Keisuke Kinoshita, Tomohiro Nakatani, Shoko Araki, Atsunori Ogawa, Takaaki Hori +8 others
  • 2011
In this paper, we introduce a system for recognizing speech in the presence of multiple rapidly time-varying noise sources. The main components of the proposed approach are a model-based speech enhancement pre-processor and an adaptation technique to optimize the integration between the pre-processor and the recognizer. The speech enhancement pre-processor(More)