Sequence Teacher-Student Training of Acoustic Models for Automatic Free Speaking Language Assessment

Conference Object
Change log
Wang, Y 
Wong, JHM 
Gales, MJF 
Knill, KM 
Ragni, A 

A high performance automatic speech recognition (ASR) system is an important constituent component of an automatic language assessment system for free speaking language tests. The ASR system is required to be capable of recognising non-native spontaneous English speech and to be deployable under real-time conditions. The performance of ASR systems can often be significantly improved by leveraging upon multiple systems that are complementary, such as an ensemble. Ensemble methods, however, can be computationally expensive, often requiring multiple decoding runs, which makes them impractical for deployment. In this paper, a lattice-free implementation of sequence-level teacher-student training is used to reduce this computational cost, thereby allowing for real-time applications. This method allows a single student model to emulate the performance of an ensemble of teachers, but without the need for multiple decoding runs. Adaptations of the student model to speakers from different first languages (L1s) and grades are also explored.

Automatic speech recognition, automatic spoken language assessment, lattice-free MMI, sequence teacher-student training, adaptation
Journal Title
2018 IEEE Spoken Language Technology Workshop, SLT 2018 - Proceedings
Conference Name
2018 IEEE Spoken Language Technology Workshop (SLT)
Journal ISSN
Volume Title
Cambridge Assessment (unknown)
Cambridge Assessment English