插补器：通过插补和动态规划进行序列建模
Imputer: Sequence Modelling via Imputation and Dynamic Programming
摘要：
本文介绍了输入，神经序列模型，通过插补迭代生成输出序列。输入器是一个迭代生成模型，只需要恒定数量的生成步骤，与输入或输出令牌的数量无关。输入器可以被训练为在输入和输出序列之间的所有可能的比对以及所有可能的生成顺序上近似边缘化。我们提出了一个易于处理的动态规划训练算法，它产生一个下界的对数边际似然。当应用于端到端语音识别时，输入器优于先前的非自回归模型，并达到与自回归模型竞争的结果。在LibriSpeech测试中，Imputer达到11.1 WER，优于CTC的13.0 WER和seq2seq的12.5 WER。
英文摘要：
This paper presents the Imputer, a neural sequence model that generates output sequences iteratively via imputations. The Imputer is an iterative generative model, requiring only a constant number of generation steps independent of the number of input or output tokens. The Imputer can be trained to approximately marginalize over all possible alignments between the input and output sequences, and all possible generation orders. We present a tractable dynamic programming training algorithm, which yields a lower bound on the log marginal likelihood. When applied to end-to-end speech recognition, the Imputer outperforms prior non-autoregressive models and achieves competitive results to autoregressive models. On LibriSpeech test-other, the Imputer achieves 11.1 WER, outperforming CTC at 13.0 WER and seq2seq at 12.5 WER.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
