Improving Translation Robustness with Visual Cues and Error Correction

03/12/2021
by   Zhenhao Li, et al.
10

Neural Machine Translation models are brittle to input noise. Current robustness techniques mostly adapt models to existing noisy texts, but these models generally fail when faced with unseen noise and their performance degrades on clean texts. In this paper, we introduce the idea of visual context to improve translation robustness against noisy texts. In addition, we propose a novel error correction training regime by treating error correction as an auxiliary task to further improve robustness. Experiments on English-French and English-German translation show that both multimodality and error correction training are beneficial for model robustness to known and new types of errors, while keeping the quality on clean texts.

READ FULL TEXT
POST COMMENT

Comments

There are no comments yet.

Authors

page 1

page 7

page 8

page 13

08/19/2018

Neural Machine Translation of Text from Non-Native Speakers

Neural Machine Translation (NMT) systems are known to degrade when confr...
10/14/2021

Understanding Model Robustness to User-generated Noisy Texts

Sensitivity of deep-neural models to input noise is known to be a challe...
10/07/2020

A Self-Refinement Strategy for Noise Reduction in Grammatical Error Correction

Existing approaches for grammatical error correction (GEC) largely rely ...
05/26/2019

Evaluation of basic modules for isolated spelling error correction in Polish texts

Spelling error correction is an important problem in natural language pr...
08/07/2020

Data Weighted Training Strategies for Grammatical Error Correction

Recent progress in the task of Grammatical Error Correction (GEC) has be...
05/25/2021

Empirical Error Modeling Improves Robustness of Noisy Neural Sequence Labeling

Despite recent advances, standard sequence labeling systems often fail w...
12/31/2020

Factual Error Correction of Claims

This paper introduces the task of factual error correction: performing e...
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.