STACL: Simultaneous Translation with Integrated Anticipation and Controllable Latency

By Mingbo Ma, Liang Huang, Hao Xiong, Kaibo Liu, Chuanqiang Zhang, Zhongjun He, Hairong Liu, Xing Li, Haifeng Wang. 2018

31 citations recorded by the source corpus.

Simultaneous translation, which translates sentences before they are finished, is useful in many scenarios but is notoriously difficult due to word-order differences and simultaneity requirements. We introduce a very simple yet surprisingly effective `wait-k' model trained to generate the target sentence concurrently with the source sentence, but always k words behind, for any given k. This framework seamlessly integrates anticipation and translation in a single model that involves only minor changes to the existing neural translation framework. Experiments on Chinese-to-English simultaneous translation achieve a 5-word latency with 3.4 (single-ref) BLEU points degradation in quality compared to full-sentence non-simultaneous translation. We also formulate a new latency metric that addresses deficiencies in previous ones.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Natural Language Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Natural Language Processing

Citations in the corpus

Robust Neural Machine Translation with Joint Textual and Phonetic Embedding (2018, methodology): “We projects the whole pinyin embedding space into a 2 -dimensional space using t-SNE technique (Maaten and Hinton, 2008 that pinyins with similar pronunciations are close to each other, such as ZHEN and ZHENG, JI and QI, MU and HU.”
Incremental Text to Speech for Neural Sequence-to-Sequence Models using Reinforcement Learning (2020, methodology): “The model is based on the prefix-to-prefix framework [18] and leverages a policy which maintains a fixed latency (in terms of number of words) behind the input.”
Thinking Slow about Latency Evaluation for Simultaneous Machine Translation (2019, methodology): “We adopt the notation of Ma et al. (2018), which in turn adopts a formalism popularized by Gris-som II et al. (2014), where the simultaneous MT system consists of an agent that begins with an empty source sentence, and must select between read actions that reveal source tokens for use in…”
Monotonic Infinite Lookback Attention for Simultaneous Machine Translation (2019, methodology): “Dalvi et al. (2018) and Ma et al. (2018) use fixed schedules and train their NMT systems accordingly.”