Waveform substitution techniques for recovering missing speech segments in packet voice communications

Goodman, D.J.; Jaffe, O.; Lockhart, G.B.; Wong, W.C.

doi:10.1109/icassp.1986.1169121

Cited by 26 publications

(24 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…If the contents of neighboring packets are correlated, the erroneous packets can be approximately reconstructed by using suitable interpolation techniques. The simplest interpolation techniques employ signal repetition [1] and signal stretching [2]. More advanced interpolation techniques are based on filter bank methods such as GAPES and MAPES [3], [4] or based on signal models such as autoregressive models [5], [6], hidden Markov models [7], and sinusoidal models [8]- [10].…”

Section: Bayesian Interpolation and Parameter Estimation In A Dynamicmentioning

confidence: 99%

“…In this paper, we initially consider the dampened sinusoidal signal model in its real form given by (1) where the sampling indices label the uniform sampled data. In the model, , ,…”

Section: Bayesian Interpolation and Parameter Estimation In A Dynamicmentioning

confidence: 99%

“…The model in (1) is also sometimes written in its polar form given by (2) where and are the undampened amplitude and phase of the th sinusoid, respectively. In this paper, we refer to the models in (1) and (2) as static sinusoidal models. This naming convention is adopted in order to distinguish it from the dynamic sinusoidal model, which we introduce later.…”

Section: Bayesian Interpolation and Parameter Estimation In A Dynamicmentioning

confidence: 99%

See 2 more Smart Citations

Bayesian Interpolation and Parameter Estimation in a Dynamic Sinusoidal Model

Nielsen

Christensen

Cemgil

et al. 2011

IEEE Trans. Audio Speech Lang. Process.

View full text Add to dashboard Cite

Abstract-In this paper, we propose a method for restoring the missing or corrupted observations of nonstationary sinusoidal signals which are often encountered in music and speech applications. To model nonstationary signals, we use a time-varying sinusoidal model which is obtained by extending the static sinusoidal model into a dynamic sinusoidal model. In this model, the in-phase and quadrature components of the sinusoids are modeled as first-order Gauss-Markov processes. The inference scheme for the model parameters and missing observations is formulated in a Bayesian framework and is based on a Markov chain Monte Carlo method known as Gibbs sampler. We focus on the parameter estimation in the dynamic sinusoidal model since this constitutes the core of model-based interpolation. In the simulations, we first investigate the applicability of the model and then demonstrate the inference scheme by applying it to the restoration of lost audio packets on a packet-based network. The results show that the proposed method is a reasonable inference scheme for estimating unknown signal parameters and interpolating gaps consisting of missing/corrupted signal segments.Index Terms-Bayesian signal processing, sinusoidal signal model, state space modeling.

show abstract

Section: Bayesian Interpolation and Parameter Estimation In A Dynamicmentioning

confidence: 99%

“…In this paper, we initially consider the dampened sinusoidal signal model in its real form given by (1) where the sampling indices label the uniform sampled data. In the model, , ,…”

Section: Bayesian Interpolation and Parameter Estimation In A Dynamicmentioning

confidence: 99%

Section: Bayesian Interpolation and Parameter Estimation In A Dynamicmentioning

confidence: 99%

See 1 more Smart Citation

Bayesian Interpolation and Parameter Estimation in a Dynamic Sinusoidal Model

Nielsen

Christensen

Cemgil

et al. 2011

IEEE Trans. Audio Speech Lang. Process.

View full text Add to dashboard Cite

show abstract

“…This ensures that the replacement will follow the changing characteristics of the whole voice stream. Studies have shown that interpolation recovery using the waveform characteristics of the sound before and after the lost packet gives sound quality better than that achieved using silence substitution or packet repetition [4]. There are variations in interpolation methods, with one method concentrating more on the pitch of the voice signal, and another on the timescale.…”

Section: Packet Lossmentioning

confidence: 99%

Internet telephony: services, technical challenges, and products

2000

View full text Add to dashboard Cite

The rapid proliferation of the Internet in the last few years has given rise to a strong interest in carrying telephony over the Internet. Because the Internet supports data communications, a range of other services can be bundled together with Internet telephony. The Internet, however, was designed for non-real-time data communications, and hence it poses several technical challenges that must be overcome before the Internet can be successfully used for carrying telephone services. This article discusses new services we can expect from Internet telephony, the technical challenges and solutions, and the emerging products that promise to support Internet telephony. INTRODUCTIONCircuit switching and packet switching are the two main technologies for computer and telecommunications networks. The current telephone system is based on circuit switching, which offers a guaranteed quality of service (QoS) to customers. A circuit has to be set up between two endpoints before the start of communication. On the contrary, data communications have so far been considered non-real-time, and hence have been carried over packet-switched networks such as the Internet, which is based on the TCP/IP protocol suite.The proliferation of the Internet in the last few years has given rise to a strong interest in the possibility of carrying real-time voice traffic over the Internet. Since the Internet was not initially designed for real-time communications, carrying voice over the Internet presents a number of challenges and technical issues that need to be solved prior to successful deployment of telephony over the Internet. Some of these technical challenges include the lack of guarantee in terms of bandwidth, packet loss, delay, and jitter which affect the quality of voice over the Internet.However, telephony over the Internet brings in a number of services and their integration which are not possible using traditional circuitswitched telephone networks. It allows the integration of voice, fax, and data over the same network, transport of intracompany voice communications between remote sites of an enterprise, and others. Services such as Web-based call centers, real-time billing, remote teleworking, and enhanced teleconferencing using shared whiteboard and shared applications are some of the services possible using IP telephony.A number of products have been developed by commercial vendors which allow voice communications over the Internet. These products can be divided into three groups according to the end user: carrier, enterprise, and individual users. The objective of this article is to bring together a survey of the new services possible with IP telephony, the technical challenges of IP telephony, and products that facilitate IP telephony. The above topics will be described in detail. APPLICATIONS AND SERVICESA major difference between the conventional public switched telephone network (PSTN) and IP telephony is that IP telephony is a voice service built on top of existing data communications services. As a result, IP te...

show abstract

“…Among several possible solutions, it is worth to mention those algorithms that try to reconstruct the missing segment of the signal from correctly received samples. For instance, waveform substitution is a method which replaces the missing part of the signal with samples of the same value as its past or future neighbours, while the pattern matching method builds a pattern from the last M known samples and searches over a window of size N the set of M samples which best matches the pattern (Goodman et al, 1986), (Tang, 1991). In (Aoki, 2004) the proposed reconstruction technique takes account of pitch variation between the previous and the next known signal frames.…”

Section: Introductionmentioning

confidence: 99%