CoDERT: Distilling Encoder Representations with Co-learning for Transducer-based Speech Recognition

We propose a simple yet effective method to compress an RNN-Transducer (RNN-T) through the well-known knowledge distillation paradigm. We show that the transducer's encoder outputs naturally have a high entropy and contain rich information about acoustically similar word-piece confusions. This rich information is suppressed when combined with the lower entropy decoder outputs to produce the joint network logits. Consequently, we introduce an auxiliary loss to distill the encoder logits from a teacher transducer's encoder, and explore training strategies where this encoder distillation works effectively. We find that tandem training of teacher and student encoders with an inplace encoder distillation outperforms the use of a pre-trained and static teacher transducer. We also report an interesting phenomenon we refer to as implicit distillation, that occurs when the teacher and student encoders share the same decoder. Our experiments show 5.37-8.4 (WERR) on in-house test sets, and 5.05-6.18 sets.


page 1

page 2

page 3

page 4


Two-Pass End-to-End ASR Model Compression

Speech recognition on smart devices is challenging owing to the small me...

Multi-head Knowledge Distillation for Model Compression

Several methods of knowledge distillation have been developed for neural...

D^3ETR: Decoder Distillation for Detection Transformer

While various knowledge distillation (KD) methods in CNN-based detectors...

Robust Knowledge Distillation from RNN-T Models With Noisy Training Labels Using Full-Sum Loss

This work studies knowledge distillation (KD) and addresses its constrai...

ERNIE-Search: Bridging Cross-Encoder with Dual-Encoder via Self On-the-fly Distillation for Dense Passage Retrieval

Neural retrievers based on pre-trained language models (PLMs), such as d...

Teacher-Student Training for Robust Tacotron-based TTS

While neural end-to-end text-to-speech (TTS) is superior to conventional...

Mimic and Conquer: Heterogeneous Tree Structure Distillation for Syntactic NLP

Syntax has been shown useful for various NLP tasks, while existing work ...

Please sign up or login with your details

Forgot password? Click here to reset