Virtual Face-to-Face Pronunciations
Research suggests that language learners can perform better with added visual pronunciation instruction versus simply audio. For
example, here's one fairly recent paper
http://mambo.ucsc.edu/pdf/interspeech2008.pdf that we've found to motivate the specific ideas behind virtual face-to-face
pronunciations. If you are deeply interested in this idea, we suggest taking a look at the citations in the end of the paper
as they trace back several years of research in visual speech training and language.