Semi-supervised DNN training with word selection for ASR

semi-supervised training, DNN, word selection, granularity of confidences

The article is about semi-supervised DNN training with word selection for Automatic Speaker Recognition (ASR).

Not all the questions related to the semi-supervised training of hybrid ASR system with DNN acoustic model were already deeply investigated. In this paper, we focus on the question of the granularity of confidences (per-sentence, per-word, perframe), the question of how the data should be used (dataselection by masks, or in mini-batch SGD with confidences as weights). Then, we propose to re-tune the system with the manually transcribed data, both with the frame CE training and sMBR training. Our preferred semi-supervised recipe which is both simple and efficient is following: we select words according to the word accuracy we obtain on the development set. Such recipe, which does not rely on a grid-search of the training hyperparameter, generalized well for: Babel Vietnamese (transcribed 11h, untranscribed 74h), Babel Bengali (transcribed 11h, untranscribed 58h) and our custom Switchboard setup (transcribed 14h, untranscribed 95h). We obtained the absolute WER improvements 2.5% for Vietnamese, 2.3% for Bengali and 3.2% for Switchboard.