Improving CTC-based ASR Models with Gated Interlayer Collaboration
This work addresses a specific bottleneck in automatic speech recognition for applications requiring accurate transcription without external language models, though it is incremental in nature.
The paper tackles the problem of CTC-based ASR models lacking capacity to model conditional dependencies and textual interactions by introducing a Gated Interlayer Collaboration mechanism, which improves performance on datasets like AISHELL-1, TEDLIUM2, and AIDATATANG, outperforming strong baselines.
The CTC-based automatic speech recognition (ASR) models without the external language model usually lack the capacity to model conditional dependencies and textual interactions. In this paper, we present a Gated Interlayer Collaboration (GIC) mechanism to improve the performance of CTC-based models, which introduces textual information into the model and thus relaxes the conditional independence assumption of CTC-based models. Specifically, we consider the weighted sum of token embeddings as the textual representation for each position, where the position-specific weights are the softmax probability distribution constructed via inter-layer auxiliary CTC losses. The textual representations are then fused with acoustic features by developing a gate unit. Experiments on AISHELL-1, TEDLIUM2, and AIDATATANG corpora show that the proposed method outperforms several strong baselines.