Anticipation-Free Training for Simultaneous Machine Translation

TL;DR

This paper introduces a novel training framework for simultaneous machine translation that separates translation and reordering, using an auxiliary sorting network to improve translation quality with lower latency.

Contribution

It proposes a new end-to-end training approach that decomposes translation and reordering, avoiding external aligners and reducing hallucinations in SimulMT.

Findings

Outperforms previous methods in translation quality

Achieves lower latency in streaming translation

Reduces hallucination phenomena during inference

Abstract

Simultaneous machine translation (SimulMT) speeds up the translation process by starting to translate before the source sentence is completely available. It is difficult due to limited context and word order difference between languages. Existing methods increase latency or introduce adaptive read-write policies for SimulMT models to handle local reordering and improve translation quality. However, the long-distance reordering would make the SimulMT models learn translation mistakenly. Specifically, the model may be forced to predict target tokens when the corresponding source tokens have not been read. This leads to aggressive anticipation during inference, resulting in the hallucination phenomenon. To mitigate this problem, we propose a new framework that decompose the translation process into the monotonic translation step and the reordering step, and we model the latter by the auxiliary sorting network (ASN). The ASN rearranges the hidden states to match the order in the target language, so that the SimulMT model could learn to translate more reasonably. The entire model is optimized end-to-end and does not rely on external aligners or data. During inference, ASN is removed to achieve streaming. Experiments show the proposed framework could outperform previous methods with less latency.