Audio-visual talking face detection
Audio-visual talking face detection
复制标题
视听说话人脸检测
DOI：
10.1109/icme.2003.1221656
复制
发表时间：
2003
期刊：
影响因子：
--
通讯作者：
I. Sethi
中科院分区：
文献类型：
--
作者：
Mingkun Li;Dongge Li;N. Dimitrova;I. Sethi
谈话人脸检测对于视频会议非常重要。然而，由于捕捉设备的低分辨率，非正式的沟通方式和背景声音，说话脸的检测是困难的。本文提出了一种利用潜在语义索引方法寻找说话脸的新方法。我们在一套全面的家庭视频会议会议中测试了我们的方法，检测率非常高。我们的实验表明，LSI方法的精度在噪声环境中会优雅地下降，而不是在存在噪声的情况下简单地失败的相关方法。
Talking face detection is important for videoconferencing. However, the detection of the talking face is difficult because of the low resolution of the capturing devices, the informal style of communication and the background sounds. In this paper, we present a novel method for finding the talking face using latent semantic indexing approach. We tested our method on a comprehensive set of home video conferencing sessions with a very high detection rate. Our experiments show that the LSI method accuracy degrades gracefully in a noisy environment as opposed to the correlation method which simply fails in presence of noise.