CIF-T: A Novel CIF-based Transducer Architecture for Automatic Speech Recognition
This work addresses efficiency and performance issues in ASR models for speech recognition applications, representing an incremental improvement over existing RNN-T architectures.
The paper tackled the computational redundancy and predictor network limitations in RNN-T models for automatic speech recognition by proposing CIF-T, which replaces the RNN-T loss with a CIF mechanism, achieving state-of-the-art results on AISHELL-1 and WenetSpeech datasets with lower computational overhead.
RNN-T models are widely used in ASR, which rely on the RNN-T loss to achieve length alignment between input audio and target sequence. However, the implementation complexity and the alignment-based optimization target of RNN-T loss lead to computational redundancy and a reduced role for predictor network, respectively. In this paper, we propose a novel model named CIF-Transducer (CIF-T) which incorporates the Continuous Integrate-and-Fire (CIF) mechanism with the RNN-T model to achieve efficient alignment. In this way, the RNN-T loss is abandoned, thus bringing a computational reduction and allowing the predictor network a more significant role. We also introduce Funnel-CIF, Context Blocks, Unified Gating and Bilinear Pooling joint network, and auxiliary training strategy to further improve performance. Experiments on the 178-hour AISHELL-1 and 10000-hour WenetSpeech datasets show that CIF-T achieves state-of-the-art results with lower computational overhead compared to RNN-T models.