Erik McDermott

Pawel Swietojanski, Stefan Braun, Dogan Can et al.

This work studies the use of attention masking in transformer transducer based speech recognition for building a single configurable model for different deployment scenarios. We present a comprehensive set of experiments comparing fixed masking, where the same attention mask is applied at every frame, with chunked masking, where the attention mask for each frame is determined by chunk boundaries, in terms of recognition accuracy and latency. We then explore the use of variable masking, where the attention masks are sampled from a target distribution at training time, to build models that can work in different configurations. Finally, we investigate how a single configurable model can be used to perform both first pass streaming recognition and second pass acoustic rescoring. Experiments show that chunked masking achieves a better accuracy vs latency trade-off compared to fixed masking, both with and without FastEmit. We also show that variable masking improves the accuracy by up to 8% relative in the acoustic re-scoring scenario.

Zijin Gu, Tatiana Likhomanenko, He Bai et al. · apple-ml

Language models play a central role in automatic speech recognition (ASR), yet most methods rely on text-only models unaware of ASR error patterns. Recently, large language models (LLMs) have been applied to ASR correction, but introduce latency and hallucination concerns. We revisit ASR error correction with compact seq2seq models, trained on ASR errors from real and synthetic audio. To scale training, we construct synthetic corpora via cascaded TTS and ASR, finding that matching the diversity of realistic error distributions is key. We propose correction-first decoding, where the correction model generates candidates rescored using ASR acoustic scores. With 15x fewer parameters than LLMs, our model achieves 1.5/3.3% WER on LibriSpeech test-clean/other, outperforms LLMs, generalizes across ASR architectures (CTC, Seq2seq, Transducer) and diverse domains, and provides precise corrections in the low-error regime where LLMs struggle.

Thien Nguyen, Nathalie Tran, Liuhui Deng et al.

Code-switching describes the practice of using more than one language in the same sentence. In this study, we investigate how to optimize a neural transducer based bilingual automatic speech recognition (ASR) model for code-switching speech. Focusing on the scenario where the ASR model is trained without supervised code-switching data, we found that semi-supervised training and synthetic code-switched data can improve the bilingual ASR system on code-switching speech. We analyze how each of the neural transducer's encoders contributes towards code-switching performance by measuring encoder-specific recall values, and evaluate our English/Mandarin system on the ASCEND data set. Our final system achieves 25% mixed error rate (MER) on the ASCEND English/Mandarin code-switching test set -- reducing the MER by 2.1% absolute compared to the previous literature -- while maintaining good accuracy on the monolingual test sets.

Stefan Braun, Erik McDermott, Roger Hsiao

The neural transducer is an end-to-end model for automatic speech recognition (ASR). While the model is well-suited for streaming ASR, the training process remains challenging. During training, the memory requirements may quickly exceed the capacity of state-of-the-art GPUs, limiting batch size and sequence lengths. In this work, we analyze the time and space complexity of a typical transducer training setup. We propose a memory-efficient training method that computes the transducer loss and gradients sample by sample. We present optimizations to increase the efficiency and parallelism of the sample-wise method. In a set of thorough benchmarks, we show that our sample-wise method significantly reduces memory usage, and performs at competitive speed when compared to the default batched computation. As a highlight, we manage to compute the transducer loss and gradients for a batch size of 1024, and audio length of 40 seconds, using only 6 GB of memory.

Adnan Haider, Xingyu Na, Erik McDermott et al.

This paper introduces a novel training framework called Focused Discriminative Training (FDT) to further improve streaming word-piece end-to-end (E2E) automatic speech recognition (ASR) models trained using either CTC or an interpolation of CTC and attention-based encoder-decoder (AED) loss. The proposed approach presents a novel framework to identify and improve a model's recognition on challenging segments of an audio. Notably, this training framework is independent of hidden Markov models (HMMs) and lattices, eliminating the need for substantial decision-making regarding HMM topology, lexicon, and graph generation, as typically required in standard discriminative training approaches. Compared to additional fine-tuning with MMI or MWER loss on the encoder, FDT is shown to be more effective in achieving greater reductions in Word Error Rate (WER) on streaming models trained on LibriSpeech. Additionally, this method is shown to be effective in further improving a converged word-piece streaming E2E model trained on 600k hours of assistant and dictation dataset.

Takaaki Hori, Martin Kocour, Adnan Haider et al.

This paper presents an efficient decoding approach for end-to-end automatic speech recognition (E2E-ASR) with large language models (LLMs). Although shallow fusion is the most common approach to incorporate language models into E2E-ASR decoding, we face two practical problems with LLMs. (1) LLM inference is computationally costly. (2) There may be a vocabulary mismatch between the ASR model and the LLM. To resolve this mismatch, we need to retrain the ASR model and/or the LLM, which is at best time-consuming and in many cases not feasible. We propose "delayed fusion," which applies LLM scores to ASR hypotheses with a delay during decoding and enables easier use of pre-trained LLMs in ASR tasks. This method can reduce not only the number of hypotheses scored by the LLM but also the number of LLM inference calls. It also allows re-tokenizion of ASR hypotheses during decoding if ASR and LLM employ different tokenizations. We demonstrate that delayed fusion provides improved decoding speed and accuracy compared to shallow fusion and N-best rescoring using the LibriHeavy ASR corpus and three public LLMs, OpenLLaMA 3B & 7B and Mistral 7B.

Roger Hsiao, Liuhui Deng, Erik McDermott et al.

We propose a novel approach to optimizing a byte-level representation for end-to-end automatic speech recognition (ASR). Byte-level representation is often used by large scale multilingual ASR systems when the character set of the supported languages is large. The compactness and universality of byte-level representation allow the ASR models to use smaller output vocabularies and therefore, provide more flexibility. UTF-8 is a commonly used byte-level representation for multilingual ASR, but it is not designed to optimize machine learning tasks directly. By using auto-encoder and vector quantization, we show that we can optimize a byte-level representation for ASR and achieve better accuracy. Our proposed framework can incorporate information from different modalities, and provides an error correction mechanism. In an English/Mandarin dictation task, we show that a bilingual ASR model built with this approach can outperform UTF-8 representation by 5% relative in error rate.

Erik McDermott, Hasim Sak, Ehsan Variani

This article describes a density ratio approach to integrating external Language Models (LMs) into end-to-end models for Automatic Speech Recognition (ASR). Applied to a Recurrent Neural Network Transducer (RNN-T) ASR model trained on a given domain, a matched in-domain RNN-LM, and a target domain RNN-LM, the proposed method uses Bayes' Rule to define RNN-T posteriors for the target domain, in a manner directly analogous to the classic hybrid model for ASR based on Deep Neural Networks (DNNs) or LSTMs in the Hidden Markov Model (HMM) framework (Bourlard & Morgan, 1994). The proposed approach is evaluated in cross-domain and limited-data scenarios, for which a significant amount of target domain text data is used for LM training, but only limited (or no) {audio, transcript} training data pairs are used to train the RNN-T. Specifically, an RNN-T model trained on paired audio & transcript data from YouTube is evaluated for its ability to generalize to Voice Search data. The Density Ratio method was found to consistently outperform the dominant approach to LM and end-to-end ASR integration, Shallow Fusion.

Qian Zhang, Han Lu, Hasim Sak et al.

In this paper we present an end-to-end speech recognition model with Transformer encoders that can be used in a streaming speech recognition system. Transformer computation blocks based on self-attention are used to encode both audio and label sequences independently. The activations from both audio and label encoders are combined with a feed-forward layer to compute a probability distribution over the label space for every combination of acoustic frame position and label history. This is similar to the Recurrent Neural Network Transducer (RNN-T) model, which uses RNNs for information encoding instead of Transformer encoders. The model is trained with the RNN-T loss well-suited to streaming decoding. We present results on the LibriSpeech dataset showing that limiting the left context for self-attention in the Transformer layers makes decoding computationally tractable for streaming, with only a slight degradation in accuracy. We also show that the full attention version of our model beats the-state-of-the art accuracy on the LibriSpeech benchmarks. Our results also show that we can bridge the gap between full attention and limited attention versions of our model by attending to a limited number of future frames.