Jan 27, 2021 · In this paper, we propose a regularization method for back-translation to explicitly draw representations of sentence pairs closer in the shared ...
In this paper, we propose a regularization method for back-translation to explicitly draw representations of sentence pairs closer in the shared space. To ...
Robust Unsupervised Neural Machine Translation With Adversarial Attack ...
ieeexplore.ieee.org › iel7
Feb 4, 2021 · In this paper, we propose a regularization method for back-translation to explicitly draw representations of sentence pairs closer in the shared ...
A2R2: Robust Unsupervised Neural Machine Translation With Adversarial Attack and Regularization on Representations ; Heng Yu, Haoran Luo, Yuqi Yi, Fan Cheng.
Dec 8, 2020 · Abstract. Unsupervised neural machine translation (UNMT) has recently attracted great interest in the ma- chine translation community.
Missing: A2R2: Attack Regularization
This paper proposes a regularization method for back-translation to explicitly draw representations of sentence pairs closer in the shared space and ...
Unlike prior works on robust NMT which generally attack on input tokens or their embeddings [32] , our model generates adversarial attack on the representations ...
Duration: 0:59
Posted: Aug 16, 2021
Posted: Aug 16, 2021
Missing: Representations. | Show results with:Representations.
We propose an approach to improving the robustness of NMT models, which consists of two parts: (1) attack the translation model with adversarial source examples ...
A2R2: Robust Unsupervised Neural Machine Translation With Adversarial Attack and Regularization on Representations. IEEE Access. 2021 | Journal article.