EURASIP Journal on Audio, Speech, and Music Processing

Table 3 Word recognition accuracy [%] for each method

From: Unsupervised domain adaptation for lip reading based on cross-modal knowledge distillation

Eval. domain		Baseline	Proposed
Source		54.62	55.07
Target	No adap.	39.19	42.84
	Known class adap.	44.10	52.22
	Unknown class adap.	—	52.04

First column shows a domain of the evaluation data

Back to article page