Localization-Driven Speech Enhancement in Noisy Multi-Speaker Hospital Environments Using Deep Learning and Meta Learning

Barhoush, Mahdi; Hallawa, Ahmed; Peine, Arne; Märtin, Lukas; Schmeink, Anke

doi:10.1109/taslp.2022.3231700

Cited by 17 publications

(2 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They can provide a great deal of information about speech characteristics. Figures 6(a) and 6(b) illustrates the separation result of the desired signal in the time domain and spectrogram [21]. The noise signal of a moving car in the time domain and spectrogram is shown in Figure 6.…”

Section:  Issn: 2722-2586mentioning

confidence: 99%

See 1 more Smart Citation

In-car speech enhancement based on source separation technique

Pathrose,

Ismail

2023

IJRA

View full text Add to dashboard Cite

This work addresses the solution of localizing and enhancing hands-free speech inside the car environment. Cars have different types of sounds from outside, co-passengers dialogue and noise. To provide better-quality speech, a microphone array-based beamforming technique is used. This research work proposes the method for selected source localization, source separation, and enhancement. An estimation of the direction of arrival (DOA) to localize the signal direction and preferred direction is selected for speech enhancement. The spiral and sine-cosine algorithm (SSCA) algorithm is combined with an adaptive least mean square to adapt the system for different environments. The algorithm is implemented in hardware and tested in a real-time car environment. The results showed significant improvement in signal-to-noise ratio (SNR) of 5.2 dB and perceptual evaluation of speech quality (PESQ) of 2.3. Finally, the model is fine-tuned for the car to get better quality. The proposed technique is efficient, and results are compared with existing methods.

show abstract

Section:  Issn: 2722-2586mentioning

confidence: 99%

“…The interference and desired signal are obtained by convoluting the masking coefficients with the input signal. The separated interference and desired signal are derived by ( 20) and (21).…”

mentioning

confidence: 99%