Noise-robust blind reverberation time estimation using noise-aware time-frequency masking

  title={Noise-robust blind reverberation time estimation using noise-aware time-frequency masking},
  author={Kaitong Zheng and Chengshi Zheng and Jinqiu Sang and Yulong Zhang and Xiaodong Li},



Noise-robust reverberation time estimation using spectral decay distributions with reduced computational cost

A novel T60 estimation algorithm based on spectral decay distributions that provides robustness to additive noise for a range of realistic noise types for signal-to-noise ratios in the range 0 to 35 dB and T60s between 200 and 950 ms is described.

Performance Comparison of Algorithms for Blind Reverberation Time Estimation from Speech

All three methods are able to estimate the reverberation time to within 0:2 s forT60 0:8 s and SNR 30 dB, while increasing the noise level causes overestimation, and the relative computational speed of the three methods is assessed.

Blind estimation of reverberation time using deep neural network

  • Myungin LeeJoon-Hyuk Chang
  • Physics, Computer Science
    2016 IEEE International Conference on Network Infrastructure and Digital Content (IC-NIDC)
  • 2016
The speech decay rate and its distribution for each frequency bin as input feature vectors of DNN are adopted and complex relation between each input feature vector and each T60 target label through multiple nonlinear hidden layers is introduced.

Blind estimators for reverberation time and direct-to-reverberant energy ratio using subband speech decomposition

Algorithms for estimating the reverberation time and direct-to-reverberant energy ratio are described, indicating the effectiveness of both techniques particularly in high-SNR situations.

Blind estimation of reverberation time.

A method for estimating RT without prior knowledge of sound sources or room geometry is presented, and results obtained for simulated and real room data are in good agreement with the real RT values.

Blind Reverberation Time Estimation Using a Convolutional Neural Network

  • H. GamperI. Tashev
  • Physics
    2018 16th International Workshop on Acoustic Signal Enhancement (IWAENC)
  • 2018
Evaluation on the ACE Challenge data corpus suggests that the proposed method is computationally efficient and outperforms state-of-the-art methods.

Robust Speaker Localization Guided by Deep Learning-Based Time-Frequency Masking

Deep learning-based time-frequency (T-F) masking has dramatically advanced monaural (single-channel) speech separation and enhancement. This study investigates its potential for direction of arrival

Blind Room Parameter Estimation Using Multiple Multichannel Speech Recordings

Results on both simulated and real data show that using multiple observations in one room significantly reduces estimation errors and variances on all target quantities, and that using two channels helps the estimation of surface and volume.

Evaluating the Non-Intrusive Room Acoustics Algorithm with the ACE Challenge

The method extracts a number of features from reverberant speech and builds a model using a recurrent neural network to estimate the reverberant acoustic parameters and finds the best method to estimate DRR provides a Root Mean Square Deviation (RMSD) of 3.84 dB.

Joint Estimation of Reverberation Time and Direct-to-Reverberation Ratio from Speech using Auditory-Inspired Features

A novel approach is proposed for joint estimation of wideband speech in noisy conditions from reverberant speech signals provided by the Acoustic Characterization of Environments (ACE) Challenge, which outperforms the baseline systems with median errors and calculation of estimates is 5.8 times faster compared to the baseline.