Multimodal Speaker Adaptation of Acoustic Model and Language Model for Asr Using Speaker Face Embedding | IEEE Conference Publication | IEEE Xplore