Distant speech separation using predicted time-frequency masks from spatial features

Pasi Pertilä, Joonas Nikunen

    Research output: Contribution to journalArticleScientificpeer-review

    29 Citations (Scopus)

    Abstract

    Speech separation algorithms are faced with a difficult task of producing high degree of separation without containing unwanted artifacts. The time-frequency (T-F) masking technique applies a real-valued (or binary) mask on top of the signal's spectrum to filter out unwanted components. The practical difficulty lies in the mask estimation. Often, using efficient masks engineered for separation performance leads to presence of unwanted musical noise artifacts in the separated signal. This lowers the perceptual quality and intelligibility of the output. Microphone arrays have been long studied for processing of distant speech. This work uses a feed-forward neural network for mapping microphone array's spatial features into a T-F mask. Wiener filter is used as a desired mask for training the neural network using speech examples in simulated setting. The T-F masks predicted by the neural network are combined to obtain an enhanced separation mask that exploits the information regarding interference between all sources. The final mask is applied to the delay-and-sum beamformer (DSB) output. The algorithm's objective separation capability in conjunction with the separated speech intelligibility is tested with recorded speech from distant talkers in two rooms from two distances. The results show improvement in instrumental measure for intelligibility and frequency-weighted SNR over complex-valued non-negative matrix factorization (CNMF) source separation approach, spatial sound source separation, and conventional beamforming methods such as the DSB and minimum variance distortionless response (MVDR).

    Original languageEnglish
    Pages (from-to)97-106
    Number of pages10
    JournalSpeech Communication
    Volume68
    DOIs
    Publication statusPublished - 2015
    Publication typeA1 Journal article-refereed

    Keywords

    • Beamforming
    • Microphone arrays
    • Neural networks
    • Speech separation
    • Time-frequency masking

    Publication forum classification

    • Publication forum level 2

    ASJC Scopus subject areas

    • Modelling and Simulation
    • Computer Science Applications
    • Computer Vision and Pattern Recognition
    • Software
    • Communication
    • Linguistics and Language
    • Language and Linguistics

    Fingerprint

    Dive into the research topics of 'Distant speech separation using predicted time-frequency masks from spatial features'. Together they form a unique fingerprint.

    Cite this