@inproceedings{broscheit-2018-learning,
    title = "Learning Distributional Token Representations from Visual Features",
    author = "Broscheit, Samuel",
    editor = "Augenstein, Isabelle  and
      Cao, Kris  and
      He, He  and
      Hill, Felix  and
      Gella, Spandana  and
      Kiros, Jamie  and
      Mei, Hongyuan  and
      Misra, Dipendra",
    booktitle = "Proceedings of the Third Workshop on Representation Learning for {NLP}",
    month = jul,
    year = "2018",
    address = "Melbourne, Australia",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/fix-sig-urls/W18-3025/",
    doi = "10.18653/v1/W18-3025",
    pages = "187--194",
    abstract = "In this study, we compare token representations constructed from visual features (i.e., pixels) with standard lookup-based embeddings. Our goal is to gain insight about the challenges of encoding a text representation from low-level features, e.g. from characters or pixels. We focus on Chinese, which{---}as a logographic language{---}has properties that make a representation via visual features challenging and interesting. To train and evaluate different models for the token representation, we chose the task of character-based neural machine translation (NMT) from Chinese to English. We found that a token representation computed only from visual features can achieve competitive results to lookup embeddings. However, we also show different strengths and weaknesses in the models' performance in a part-of-speech tagging task and also a semantic similarity task. In summary, we show that it is possible to achieve a \textit{text representation} only from pixels. We hope that this is a useful stepping stone for future studies that exclusively rely on visual input, or aim at exploiting visual features of written language."
}