Results 151 to 160 of about 1,864,117 (195)
Modeling audio dynamics using hierarchical assisted K-means model for structured speaker profiling in TED talks. [PDF]
Sci RepRenukadevi MN +6 more
europepmc +1 more source
Continuous Speech Separation With Conformer [PDF]
ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2021 Continuous speech separation plays a vital role in complicated speech related tasks such as conversation transcription. The separation model extracts a single speaker signal from a mixed speech. In this paper, we use transformer and conformer in lieu of recurrent neural networks in the separation system, as we believe capturing global information with ...
Jinyu Li, Sanyuan Chen, Takuya Yoshioka
exaly +4 more sources
Continuous Speech Separation: Dataset and Analysis [PDF]
ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2020 This paper describes a dataset and protocols for evaluating continuous speech separation algorithms. Most prior studies on speech separation use pre-segmented signals of artificially mixed speech utterances which are mostly \emph{fully} overlapped, and the algorithms are evaluated based on signal-to-distortion ratio or similar performance metrics ...
Liang Lu, Jinyu Li, Takuya Yoshioka
exaly +4 more sources
Some of the next articles are maybe not open access.
Related searches:
Related searches:
ICASSP '87. IEEE International Conference on Acoustics, Speech, and Signal Processing, 1988
The multisignal minimum-cross-entropy spectral analysis (multisignal MCESA) is applied to the problem of separating the speech signals of two talkers speaking simultaneously on a single channel, e.g., when two talkers use a single microphone. A new two-stage approach to the problem is proposed in which a spectral separator is followed by a spectral ...
D. G. Childers, C. K. Lee
openaire +2 more sources
The multisignal minimum-cross-entropy spectral analysis (multisignal MCESA) is applied to the problem of separating the speech signals of two talkers speaking simultaneously on a single channel, e.g., when two talkers use a single microphone. A new two-stage approach to the problem is proposed in which a spectral separator is followed by a spectral ...
D. G. Childers, C. K. Lee
openaire +2 more sources
2008
The PASCAL Speech Separation Challenge (SSC) is based on a corpus of sentences from the Wall Street Journal task read by two speakers simultaneously and captured with two circular eight-channel microphone arrays. This work describes our system for the recognition of such simultaneous speech.
John W. McDonough +7 more
openaire +1 more source
The PASCAL Speech Separation Challenge (SSC) is based on a corpus of sentences from the Wall Street Journal task read by two speakers simultaneously and captured with two circular eight-channel microphone arrays. This work describes our system for the recognition of such simultaneous speech.
John W. McDonough +7 more
openaire +1 more source
2010
The work of Bernstein and Benoit has confirmed that it is advantageous to use multiple senses, for example to employ both audio and visual modalities, in speech perception. As a consequence, looking at the speaker's face can be useful to better hear a speech signal in a noisy environment and to extract it from competing sources, as originally ...
Rivet, B, Chambers, J
openaire +4 more sources
The work of Bernstein and Benoit has confirmed that it is advantageous to use multiple senses, for example to employ both audio and visual modalities, in speech perception. As a consequence, looking at the speaker's face can be useful to better hear a speech signal in a noisy environment and to extract it from competing sources, as originally ...
Rivet, B, Chambers, J
openaire +4 more sources
Phoneme-specific speech separation
2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2016Speech separation or enhancement algorithms seldom exploit information about phoneme identities. In this study, we propose a novel phoneme-specific speech separation method. Rather than training a single global model to enhance all the frames, we train a separate model for each phoneme to process its corresponding frames.
Zhong-Qiu Wang 0001 +2 more
openaire +2 more sources
Speech separation by efficient combinatorial decoding of speech mixtures
2009 IEEE International Conference on Multimedia and Expo, 2009We formulate the cocktail party problem as the minimization of a symmetric posimodular function defined on fragments of the signal captured by a single microphone. This formulation allows the application of tractable combinatorial optimization techniques, and in particular the Queyranne's algorithm [1], to exactly solve a problem which was previously ...
Manuel Reyes-Gomez, Nebojsa Jojic
openaire +2 more sources