English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Simultaneous Machine Translation (SiMT) aims to yield a real-time partialtranslation with a monotonically growing the source-side context. However,there is a counterintuitive phenomenon about the context usage between trainingand testing: e.g., the wait-k testing model consistently trained with wait-k ismuch worse than that model inconsistently trained with wait-k' (k' is not equalto k) in terms of translation quality. To this end, we first investigate theunderlying reasons behind this phenomenon and uncover the following twofactors: 1) the limited correlation between translation quality and training(cross-entropy) loss; 2) exposure bias between training and testing. Based onboth reasons, we then propose an effective training approach called contextconsistency training accordingly, which makes consistent the context usagebetween training and testing by optimizing translation quality and latency asbi-objectives and exposing the predictions to the model during the training.The experiments on three language pairs demonstrate our intuition: our systemencouraging context consistency outperforms that existing systems with contextinconsistency for the first time, with the help of our context consistencytraining approach.

Context Consistency between Training and Testing in Simultaneous Machine Translation