We introduce a Multi-modal Neural Machine Translation model in which a doubly-attentive decoder naturally incorporates spatial visual features obtained using pre-trained convolutional neural networks, bridging the gap between image description and translation. Our decoder learns to attend to source-language words and parts of an image independently by means of two separate attention mechanisms as it generates words in the target language. We find that our model can efficiently exploit not just back-translated in-domain multi-modal data but also large general-domain text-only MT corpora. We also report state-of-the-art results on the Multi30k data set.

This paper discusses neural machine translation (NMT), a new paradigm in the MT field, comparing the quality of NMT systems with statistical MT by describing three studies using automatic and human evaluation methods. Automatic evaluation results presented for NMT are very promising, however human evaluations show mixed results. We report increases in fluency but inconsistent results for adequacy and post-editing effort. NMT undoubtedly represents a step forward for the MT field, but one that the community should be careful not to oversell.

We introduce multi-modal, attentionbased Neural Machine Translation (NMT) models which incorporate visual features into different parts of both the encoder and the decoder. Global image features are extracted using a pre-trained convolutional neural network and are incorporated (i) as words in the source sentence, (ii) to initialise the encoder hidden state, and (iii) as additional data to initialise the decoder hidden state. In our experiments, we evaluate translations into English and German, how different strategies to incorporate global image features compare and which ones perform best. We also study the impact that adding synthetic multi-modal, multilingual data brings and find that the additional data have a positive impact on multi-modal models. We report new state-of-the-art results and our best models also significantly improve on a comparable Phrase-Based Statistical MT (PBSMT) model trained on the Multi30k data set according to all metrics evaluated. To the best of our knowledge, it is the first time a purely neural model significantly improves over a PBSMT model on all metrics evaluated on this data set.

We present a doubly-attentive multimodal machine translation model. Our model learns to attend to source language and spatial-preserving CONV 5,4 visual features as separate attention mechanisms in a neural translation model. In image description translation experiments (Task 1), we find an improvement of 2.3 Meteor points compared to initialising the hidden state of the decoder with only the FC 7 features and 2.9 Meteor points compared to a text-only neural machine translation baseline, confirming the useful nature of attending to the CONV 5,4 features.

In this work, we propose to model the interaction between visual and textual features for multi-modal neural machine translation (MMT) through a latent variable model. This latent variable can be seen as a multi-modal stochastic embedding of an image and its description in a foreign language. It is used in a target-language decoder and also to predict image features. Importantly, our model formulation utilises visual and textual inputs during training but does not require that images be available at test time. We show that our latent variable MMT formulation improves considerably over strong baselines, including a multi-task learning approach (Elliott and Kádár, 2017) and a conditional variational auto-encoder approach (Toyama et al., 2016). Finally, we show improvements due to (i) predicting image features in addition to only conditioning on them, (ii) imposing a constraint on the minimum amount of information encoded in the latent variable, and (iii) by training on additional target-language image descriptions (i.e. synthetic data).1 Code and pre-trained models will be released soon.