顔画像ビデオシステムにおけるオーディオ・リップリーディング
スポンサーリンク
概要
- 論文の詳細を見る
Multimedia has a large demand for both bandwidth and hardware resources (video cameras,video compression boards etc.). These present an obstacle to realisation of a usable video-conferencing system on the many 10Mb/s ethernet LANs existing today. MPEG-4 aims to provide very low bandwidth facial video at 9-40Kb/s rates, alleviating the bandwidth shortage, bet not the demand on hardwear. An alternative to compression in hardwear for face-to-face video is their reconstruction at the receiver by analysis of the received voice signal. A prerequisite for this task is continuous, fast (real-time) lip movement recognition from speech input alone, Audio Lip-reading.
- 1994-09-26
論文 | ランダム
- 最終講義 温故知新--中国近代思想史研究三十年 (竹内弘行教授 退休記念号)
- 訳注 康有為『大同書草稿』訳注(6)
- 訳注 康有為『大同書草稿』訳注(5)
- 訳注 康有為『大同書草稿』訳注(4)
- 訳注 康有為『大同書草稿』訳注