A real-time blind source separation scheme and its application to reverberant and noisy acoustic environments

Aichner, Robert; Buchner, Herbert; Yan, Fei; Kellermann, Walter

doi:10.1016/j.sigpro.2005.06.022

Cited by 78 publications

(98 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…For 4-source cases, large misalignment occurs during the clustering process, resulting in the poor separation performance. Comparing to the benchmark, all methods show excellent performance for 2 2 cases, and degraded performance for 4 4 cases. However, the proposed method is closer to the ideal separation results than other methods especially for 4 4 cases.…”

Section: B Bss Experiments In Different Simulation Conditionsmentioning

confidence: 97%

“…The first is to make the separation filters smooth in the frequency domain [4], [10], [11]. This may be achieved by limiting their lengths.…”

Section: A Region-growing Permutation Alignment Approach In Frequencymentioning

confidence: 99%

“…To evaluate the performance of the proposed permutation alignment scheme efficiently, we use the same frequency-domain BSS processing for the proposed method and methods 3-6 (Murata(Env), Murata(Pow), PowerRatio, and Benchmark): with all processing stages and parameters fixed but different permutation correction schemes. Table I shows the average output SIRs for all cases, 4 where "n/a" means the algorithm fails in the experiment condition. Note that the average input SIR is about 0 dB for the 2 2 case and 5 dB for the 4 4 case.…”

Section: B Bss Experiments In Different Simulation Conditionsmentioning

confidence: 99%

See 2 more Smart Citations

A Region-Growing Permutation Alignment Approach in Frequency-Domain Blind Source Separation of Speech Mixtures

Wang

Ding

Yin

2011

IEEE Trans. Audio Speech Lang. Process.

View full text Add to dashboard Cite

Abstract-The convolutive blind source separation (BSS) problem can be solved efficiently in the frequency domain, where instantaneous BSS is performed separately in each frequency bin. However, the permutation ambiguity in each frequency bin should be resolved so that the separated frequency components from the same source are grouped together. To solve the permutation problem, this paper presents a new alignment method based on an inter-frequency dependence measure: the powers of separated signals. Bin-wise permutation alignment is applied first across all frequency bins, using the correlation of separated signal powers; then the full frequency band is partitioned into small regions based on the bin-wise permutation alignment result. Finally, region-wise permutation alignment is performed in a region-growing manner. The region-wise permutation correction scheme minimizes the spreading of the misalignment at isolated frequency bins to others, hence to improve permutation alignment. Experiment results in simulated and real environments verify the effectiveness of the proposed method. Analysis demonstrates that the proposed frequency-domain BSS method is computationally efficient.

show abstract

Section: B Bss Experiments In Different Simulation Conditionsmentioning

confidence: 97%

“…The first is to make the separation filters smooth in the frequency domain [4], [10], [11]. This may be achieved by limiting their lengths.…”

Section: A Region-growing Permutation Alignment Approach In Frequencymentioning

confidence: 99%

Section: B Bss Experiments In Different Simulation Conditionsmentioning

confidence: 99%

See 1 more Smart Citation

A Region-Growing Permutation Alignment Approach in Frequency-Domain Blind Source Separation of Speech Mixtures

Wang

Ding

Yin

2011

IEEE Trans. Audio Speech Lang. Process.

View full text Add to dashboard Cite

show abstract

“…Convolutive BSS as applied to speech signal mixtures involves relatively-long multichannel FIR filters to achieve separation with even moderate amounts of room reverberation. While time-domain algorithms can be developed to perform this task, they can be difficult to code primarily due to the multichannel convolution operations involved [5,6]. One way to simplify the conceptualization of the convolutive BSS algorithms is to transform the task into the frequency domain, as convolution in time becomes multiplication in frequency.…”

Section: Mixing Process and Convolutive Bssmentioning

confidence: 99%

“…They can be classified into two major approaches: time-domain BSS [5,6] and frequency-domain BSS [7]. With time-domain BSS, a cost function is defined for time-domain signals, and optimized with convolutive separation filters.…”

Section: Introductionmentioning

confidence: 99%