The Influence of Cross-Modal Interaction on Audio-Visual Speech Quality Perception

Rimell AN, Hollier MP, Voelcker RM. The Influence of Cross-Modal Interaction on Audio-Visual Speech Quality Perception. In: AES Convention 105. Audio Engineering Society; 1998. Paper 4791. Available from: https://aes.org/publications/elibrary-page/?id=8389

@inproceedings{Rimell1998_8389,
  author = {Rimell, Andrew N. and Hollier, Michael P. and Voelcker, R. M.},
  title = {{The Influence of Cross-Modal Interaction on Audio-Visual Speech Quality Perception}},
  booktitle = {AES Convention 105},
  note = {Paper 4791},
  year = {1998},
  month = sep,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=8389}
}

TY  - CPAPER
TI  - The Influence of Cross-Modal Interaction on Audio-Visual Speech Quality Perception
AU  - Rimell, Andrew N.
AU  - Hollier, Michael P.
AU  - Voelcker, R. M.
T2  - AES Convention 105
M1  - Paper 4791
PY  - 1998
DA  - 1998/09/06
UR  - https://aes.org/publications/elibrary-page/?id=8389
PB  - Audio Engineering Society
LA  - en
AB  - Combined audio-visual services are expected to become a feature of the next generation of telecommunication systems. Current systems apply individual perceptually motivated data reduction to the audio and video. A system which applies perceptually motivated data reduction to the combined audio-visual content may provide greater data reduction, whilst maintaining high quality service over low bit-rate transmission media. To be able to design such bimodal codecs, and optimize overall performance, (through the use of low bit-rate codecs),it is important to understand the trade-offs between the quality of the audio and the video. This paper compares the results of a complimentary pair of subjective experiments designed to investigate the differences between visual speech and non-visual speech quality perception and quality mismatch. Conclusions are drawn about the variation in cross-modal interaction, and particularly speech quality perception, for different audio-visual content. The results obtained will be used in the design of a bi-model perceptual model.
ER  -