Energy-efficient soft real-time CPU scheduling for mobile multimedia systems
SOSP '03 Proceedings of the nineteenth ACM symposium on Operating systems principles
Probabilistic Finite-State Machines-Part I
IEEE Transactions on Pattern Analysis and Machine Intelligence
Source-driven packet marking for speech transmission over differentiated-services networks
ICASSP '01 Proceedings of the Acoustics, Speech, and Signal Processing, 200. on IEEE International Conference - Volume 02
Probabilistic design of multimedia embedded systems
ACM Transactions on Embedded Computing Systems (TECS)
A semi-continuous state-transition probability HMM-based voice activity detector
EURASIP Journal on Audio, Speech, and Music Processing
| Hi-index | 0.00 |
In this paper, we describe the formatting guidelines for ACM SIG Proceedings. Traditionally silence in VoIP applications is detected by a VAD algorithm after G.729 compression, both have high computational and energy cost. However, such cost on silence frames will be completely wasted because the detected silence frames will not be sent to receiver. We propose to use a silence pre-detection (SPD) module to detect silence frames directly from the voice sample in order to reduce energy. Our adaptive SPD algorithm has little hardware requirement and low computation cost. However, it can detect 59% of the silence, which results in 34% energy saving. The impact to quality of speech is almost unnoticeable.