Joint dictionary training for bandwidth extension of speech signals

Sadasivan, Jishnu; Mukherjee, Subhadip; Seelamantula, Chandra Sekhar

doi:10.1109/icassp.2016.7472814

Cited by 14 publications

(4 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Source-filter models have been also employed to extend the bandwidth using line spectral frequencies in [15]. Systems based on dictionary learning to map low-frequency patterns to high-frequency components have been proposed in [16,17]. Classic machine learning methods have also been explored for BWE, such as Gaussian mixture models (GMMs) [18], hidden Markov models (HMM) [19,20], or non-negative matrix factorization (NMF) [21,22].…”

Section: Signal Processing Approachesmentioning

confidence: 99%

Efficient bandwidth extension of musical signals using a differentiable harmonic plus noise model

Grumiaux,

Lagrange

2023

J AUDIO SPEECH MUSIC PROC.

View full text Add to dashboard Cite

The task of bandwidth extension addresses the generation of missing high frequencies of audio signals based on knowledge of the low-frequency part of the sound. This task applies to various problems, such as audio coding or audio restoration. In this article, we focus on efficient bandwidth extension of monophonic and polyphonic musical signals using a differentiable digital signal processing (DDSP) model. Such a model is composed of a neural network part with relatively few parameters trained to infer the parameters of a differentiable digital signal processing model, which efficiently generates the output full-band audio signal. We first address bandwidth extension of monophonic signals, and then propose two methods to explicitly handle polyphonic signals. The benefits of the proposed models are first demonstrated on monophonic and polyphonic synthetic data against a baseline and a deep-learning-based ResNet model. The models are next evaluated on recorded monophonic and polyphonic data, for a wide variety of instruments and musical genres. We show that all proposed models surpass a higher complexity deep learning model for an objective metric computed in the frequency domain. A MUSHRA listening test confirms the superiority of the proposed approach in terms of perceptual quality.

show abstract

Section: Signal Processing Approachesmentioning

confidence: 99%

Efficient bandwidth extension of musical signals using a differentiable harmonic plus noise model

Grumiaux,

Lagrange

2023

J AUDIO SPEECH MUSIC PROC.

View full text Add to dashboard Cite

show abstract

“…The bandwidth extension is then divided into two separate tasks: spectral envelope extension and excitation signal extension. Early studies employ codebook mapping techniques for spectral envelope extension, using two codebooks representing narrowband and wideband spectra H. Wang is with the Department of Computer Science and Engineering, The Ohio State University, OH 43210, USA (e-mail: wang.11401@osu.edu) E. W. Healy is with the Department of Speech and Hearing Science and the Center for Cognitive and Brain Sciences, The Ohio State University, Columbus, OH 43210, USA (e-mail: healy.66@osu.edu) D. L. Wang is with the Department of Computer Science and Engineering and the Center for Cognitive and Brain Sciences, The Ohio State University, Columbus, OH 43210, USA (e-mail: dwang@cse.ohio-state.edu) [52], [58]. Through training, the best matching entry in the wideband codebook to the narrowband codebook is used to generate the spectral envelope.…”

Section: Introductionmentioning

confidence: 99%

Towards Robust Speech Super-Resolution

Wang

2021

IEEE/ACM Trans. Audio Speech Lang. Process.

View full text Add to dashboard Cite

Speech super-resolution (SR) is the task that restores high-resolution speech from low-resolution input. Existing models employ simulated data and constrained experimental settings, which limit generalization to real-world SR. Predictive models are known to perform well in fixed experimental settings, but can introduce artifacts in adverse conditions. On the other hand, generative models learn the distribution of target data and have a better capacity to perform well on unseen conditions. In this study, we propose a novel two-stage approach that combines the strengths of predictive and generative models. Specifically, we employ a diffusion-based model that is conditioned on the output of a predictive model. Our experiments demonstrate that the model significantly outperforms single-stage counterparts and existing strong baselines on benchmark SR datasets. Furthermore, we introduce a repainting technique during the inference of the diffusion process, enabling the proposed model to regenerate high-frequency components even in mismatched conditions. An additional contribution is the collection of and evaluation on real SR recordings, using the same microphone at different native sampling rates. We make this dataset freely accessible, to accelerate progress towards real-world speech super-resolution.

show abstract

“…In Kim et al [2008], temporal envelope and fine structure of sub-bands are taken into account for ABE. In Li and Lee [2015], Sadasivan et al [2016], Bin et al [2015], Bachhav et al [2017], the WB magnitude spectrum is taken directly for extracting bene-ficial HB and NB information. In Sunil and Sinha [2014], sparse modeling is used for ABE, wherein it requires a separate dictionary for voiced speech and unvoiced speech.…”

Section: Introductionmentioning

confidence: 99%

Artificial bandwidth extension using deep neural network and $H^\infty$ sampled-data control theory

Gupta,

Shekhawat

2021

Preprint

View full text Add to dashboard Cite

Artificial bandwidth extension is applied to speech signals to improve their quality in narrowband telephonic communication. For accomplishing this, the missing high-frequency (high-band) components of speech signals are recovered by utilizing a new extrapolation process based on sampleddata control theory and deep neural network (DNN). The H ∞ sampled-data control theory helps in designing of a high-band filter to recover the high-frequency signals by optimally utilizing the inter-sample signals. Non-stationary (time-varying) characteristics of speech signals forces to use numerous high-band filters. Hence, we use a deep neural network for estimating the high-band filter information and a gain factor for a specified narrowband information of the unseen signal. The objective analysis is done on the TIMIT dataset and RSR15 dataset. Additionally, the objective analysis is performed separately for the voiced speech as well as for the unvoiced speech as generally needed in speech processing. Subjective analysis is done on the RSR15 dataset.

show abstract

Joint dictionary training for bandwidth extension of speech signals

Cited by 14 publications

References 23 publications

Efficient bandwidth extension of musical signals using a differentiable harmonic plus noise model

Efficient bandwidth extension of musical signals using a differentiable harmonic plus noise model

Towards Robust Speech Super-Resolution

Artificial bandwidth extension using deep neural network and $H^\infty$ sampled-data control theory

Contact Info

Product

Resources

About