scispace - formally typeset
Search or ask a question
Topic

Voice activity detection

About: Voice activity detection is a research topic. Over the lifetime, 12784 publications have been published within this topic receiving 272632 citations. The topic is also known as: speech activity detection & speech detection.


Papers
More filters
01 Jan 2019
TL;DR: This paper proposes the extraction of hand-crafted appearance and optical flow features from the video modality, and time-domain Features from the accelerometer data, and shows that applying a multimodal late fusion technique can lead to a performance boost in most cases.
Abstract: In this paper we examine the task of automatic detection of speech without microphones, using an overhead camera and wearable accelerometers. For this purpose, we propose the extraction of hand-crafted appearance and optical flow features from the video modality, and time-domain features from the accelerometer data. We evaluate the performance of the separate modalities in a large dataset of over 25 hours of standing conversation between multiple individuals. Finally, we show that applying a multimodal late fusion technique can lead to a performance boost in most cases.

1 citations

Patent
27 Dec 2013
TL;DR: In this paper, the authors present improvements for speech recognition systems used to control devices, such as two-stage confirmation, limited speech recognition, and wake-up for speech driven applications and systems.
Abstract: Presented are improvements for speech recognition systems used to control devices. Features include two-stage confirmation, two-stage limited speech recognition mode, and two-stage wake-up for speech driven applications and systems. A headset computer device includes such staged confirmation operation.

1 citations

01 Jan 2013
TL;DR: This review paper presented the basic of spectral subtraction algorithm, minimum mean square error, wiener algorithm and TSDD algorithm, and performance evaluation of various modified decision-directed approach.
Abstract: In this paper, the aim of speech enhancement algorithms is to improve the quality or intelligibility of the noisy speech signals by using different enhancement algorithms. Many speech enhancement algorithms are designed to suppress additive background noise. This review paper presented the basic of spectral subtraction algorithm, minimum mean square error, wiener algorithm and TSDD algorithm, and performance evaluation of various modified decision-directed approach. This paper provides valuable hints for analyzing and optimizing noise- reduction algorithm. The speech enhancement methods aimed at suppressing the background noise are based on one way or the other on the estimation of the background noise. If the background noise is evolving more slowly than the speech, i.e., if the noise is more stationary than the speech, it is easy to estimate the noise during the pauses in speech. This paper also reports the subjective and objective tests. Keywords: speech distortions, speech enhancement algorithm, and speech intelligibility improvement.

1 citations

Proceedings ArticleDOI
12 Apr 1994
TL;DR: In this work a speech synthesis system that uses concatenation of phoneme waveforms as the method of synthesis increases the intelligibility of the speech but also increases the storage needs of the system.
Abstract: In this work a speech synthesis system is implemented. The system uses concatenation of phoneme waveforms as the method of synthesis. These waveforms are generated by sampling the speech of a human speaker and then separating it into its phonemes. These phoneme samples are stored in the hard disk to be used in the synthesis. Then the text to be read is separated into its syllables and each syllable is synthesized by concatenating the phoneme samples. This method is facilitated by the structure of the Turkish language and some exceptions are taken into account. The same synthesis method is then applied using diphones as the units of synthesis. This increases the intelligibility of the speech but also increases the storage needs of the system. >

1 citations


Network Information
Related Topics (5)
Signal processing
73.4K papers, 983.5K citations
81% related
Feature vector
48.8K papers, 954.4K citations
80% related
Decoding methods
65.7K papers, 900K citations
79% related
Recurrent neural network
29.2K papers, 890K citations
79% related
Feature extraction
111.8K papers, 2.1M citations
78% related
Performance
Metrics
No. of papers in the topic in previous years
YearPapers
2023121
2022266
2021301
2020300
2019262
2018238