A wake-up method, an apparatus, a device, and a computer storage medium, which are used in a voice device. The method comprises: obtaining audio to be identified (S101); utilizing a wake-up model and at least two groups of training data to separately perform processing on the audio to be identified, and obtaining at least two confidence scores and a confidence score threshold corresponding to each score (S 102), wherein the at least two groups of training data are obtained by separately performing training on at least two wake-up word training sets by means of the wake-up model; and triggering a wake-up event of the voice device according to a comparison result between the at least two confidence scores and the confidence score threshold corresponding to each score (S 103).