Skip to main content

Research Repository

Advanced Search

Variational Recurrent sequence to sequence retrieval for stepwise illustration

He, Y; Vogiatzis, G; Guha, T; Ferhatosmanoglu, H; Batra, V; Haldar, A

Variational Recurrent sequence to sequence retrieval for stepwise illustration Thumbnail


Authors

Y He

G Vogiatzis

T Guha

H Ferhatosmanoglu

A Haldar



Abstract

We address and formalise the task of sequence-to-sequence (seq2seq) cross-modal retrieval. Given a sequence of text passages as query, the goal is to retrieve a sequence of images that best describes and aligns with the query. This new task extends the traditional cross-modal retrieval, where each image-text pair is treated independently ignoring broader context. We propose a novel variational recurrent seq2seq (VRSS) retrieval model for this seq2seq task. Unlike most cross-modal methods, we generate an image vector corresponding to the latent topic obtained from combining the text semantics and context. This synthetic image embedding point associated with every text embedding point can then be employed for either image generation or image retrieval as desired. We evaluate the model for the application of stepwise illustration of recipes, where a sequence of relevant images are retrieved to best match the steps described in the text. To this end, we build and release a new Stepwise Recipe dataset for research purposes, containing 10K recipes (sequences of image-text pairs) having a total of 67K image-text pairs. To our knowledge, it is the first publicly available dataset to offer rich semantic descriptions in a focused category such as food or recipes. Our model is shown to outperform several competitive and relevant baselines in the experiments. We also provide qualitative analysis of how semantically meaningful the results produced by our model are through human evaluation and comparison with relevant existing methods.

Citation

He, Y., Vogiatzis, G., Guha, T., Ferhatosmanoglu, H., Batra, V., & Haldar, A. (2020). Variational Recurrent sequence to sequence retrieval for stepwise illustration. https://doi.org/10.1007/978-3-030-45439-5_4

Acceptance Date Dec 9, 2019
Publication Date Apr 8, 2020
Journal Proceedings ofAdvances in Information Retrieval ECIR 2020
Pages 50-64
Series Title European Conference on Information Retrieval (ECIR)
DOI https://doi.org/10.1007/978-3-030-45439-5_4
Keywords Multimodal datasets, Semantics, Sequence retrieval
Publisher URL https://link.springer.com/chapter/10.1007%2F978-3-030-45439-5_4

Files




You might also like



Downloadable Citations