Zero-Shot Translation using Diffusion Models

By Eliya Nachmani, Shaked Dovrat. 2021

13 citations recorded by the source corpus.

In this work, we show a novel method for neural machine translation (NMT), using a denoising diffusion probabilistic model (DDPM), adjusted for textual data, following recent advances in the field. We show that it's possible to translate sentences non-autoregressively using a diffusion model conditioned on the source sentence. We also show that our model is able to translate between pairs of languages unseen during training (zero-shot learning).

Subfields: Artificial Intelligence, Generative Models, Machine Translation

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

1. Jun Zhu
2. Chongxuan Li
3. Hang Su
4. Fan Bao
5. Samuli Laine
6. Timo Aila
7. Miika Aittala
8. Kaiwen Xue

In Generative Models

1. Tero Karras
2. Timo Aila
3. Richard Zhang
4. Eli Shechtman
5. Jun-Yan Zhu
6. Samuli Laine
7. Miika Aittala
8. Daniel Cohen-Or

Citations in the corpus

• Elucidating the Design Space of Diffusion-Based Generative Models (2022, background): “They are also rapidly finding use in other domains such as audio [27, 37] and video [19] generation, image segmentation [4, 54] and language translation [34].”
• All are Worth Words: A ViT Backbone for Diffusion Models (2022, methodology): “In addition to images, transformers in diffusion models are also employed to encode texts [44,47,49,51], decode texts [7, 28, 36, 43] and generate CLIP embeddings [47].”