Exploring Neural Transducers for End-to-End Speech Recognition
This work addresses the problem of simplifying speech recognition pipelines for researchers and practitioners by showing competitive performance without language models, though it is incremental as it compares existing methods.
The authors compared CTC, RNN-Transducer, and attention-based Seq2Seq models for end-to-end speech recognition, finding that Seq2Seq and RNN-Transducer outperform CTC with a language model on the Hub5'00 benchmark and an internal dataset, simplifying the pipeline to neural network operations.
In this work, we perform an empirical comparison among the CTC, RNN-Transducer, and attention-based Seq2Seq models for end-to-end speech recognition. We show that, without any language model, Seq2Seq and RNN-Transducer models both outperform the best reported CTC models with a language model, on the popular Hub5'00 benchmark. On our internal diverse dataset, these trends continue - RNNTransducer models rescored with a language model after beam search outperform our best CTC models. These results simplify the speech recognition pipeline so that decoding can now be expressed purely as neural network operations. We also study how the choice of encoder architecture affects the performance of the three models - when all encoder layers are forward only, and when encoders downsample the input representation aggressively.