Counting competing speakers in a timeframe — human versus computer

Andrei, Valentin; Cucu, Horia; Buzo, Andi; Burileanu, Corneliu

doi:10.21437/interspeech.2015-673

Cited by 5 publications

(4 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since we are dealing with a novel task description, related speaker count estimation techniques like those introduced in Section 1, could hardly be used as baselines. Specifically, [32] does not work on fully overlapped speech, [2] does not scale to the size of our dataset, since it requires to crosscorrelate the full database against another. Finally, [25] proposes a feature but does not employ a fully automated system.…”

Section: Discussionmentioning

confidence: 99%

“…In another vein, Andrei et.al. [2] proposed an algorithm which correlates single frames of multi-speaker mixtures with a set of single-speaker utterances. Motivated by the recent and impressive successes of deep learning approaches in various audio-related tasks [13,33], we focus on developing such a method for direct count estimation.…”

Section: Introductionmentioning

confidence: 99%

“…For the sake of reproducibility, pre-trained network and the test dataset are made available for download on the accompanying website. 2…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Classification vs. Regression in Supervised Learning for Single Channel Speaker Count Estimation

Stöter

Chakrabarty

Edler

et al. 2018

2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

The task of estimating the maximum number of concurrent speakers from single channel mixtures is important for various audio-based applications, such as blind source separation, speaker diarisation, audio surveillance or auditory scene classification. Building upon powerful machine learning methodology, we develop a Deep Neural Network (DNN) that estimates a speaker count. While DNNs efficiently map input representations to output targets, it remains unclear how to best handle the network output to infer integer source count estimates, as a discrete count estimate can either be tackled as a regression or a classification problem. In this paper, we investigate this important design decision and also address complementary parameter choices such as the input representation. We evaluate a stateof-the-art DNN audio model based on a Bi-directional Long Short-Term Memory network architecture for speaker count estimations. Through experimental evaluations aimed at identifying the best overall strategy for the task and show results for five seconds speech segments in mixtures of up to ten speakers.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Classification vs. Regression in Supervised Learning for Single Channel Speaker Count Estimation

Stöter

Chakrabarty

Edler

et al. 2018

2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

show abstract

“…Detecting overlapped speech on short timeframes can contribute to key BSS applications. In some previous papers [13], [14] we presented several methods for competing speaker counting and compared them with the human capabilities of counting the speakers in an overlapped speech recorded on a single channel.…”

Section: Detecting Overlapped Speech On Short Timeframes Using Deep L...mentioning

confidence: 99%

Machine Learning for Spoken Language Technology

Burileanu¹,

Mihalache²,

Andrei³

et al. 2021

AnnalsARSciInfo

Self Cite

View full text Add to dashboard Cite

Spoken language technology is one of the domains in which, in our days, machine learning algorithms and especially neural networks are used. Some applications will pe presented in this paper: detecting overlapped speech on short time frames (till 25ms), emotion recognition from speech (including speech stress detection and deceptive speech detection) and the performances of the last large vocabulary continuous speech recognition systems for Romanian developed in the SpeeD Laboratory, from Research Institute “CAMPUS”, University POLITEHNICA of Bucharest

show abstract