ML Anthology

Improved Adversarial Image Captioning

Pierre Dognin, Igor Melnyk, Youssef Mroueh, Jarret Ross, Tom Sercu
ICLRW 2019
/iclrw/2019/dognin2019iclrw-improved/

Abstract

In this paper we study image captioning as a conditional GAN training, proposing both a context-aware LSTM captioner and co-attentive discriminator, which enforces semantic alignment between images and captions. We investigate the viability of two discrete GAN training methods: Self-critical Sequence Training (SCST) and Gumbel Straight-Through (ST) and demonstrate that SCST shows more stable gradient behavior and improved results over Gumbel ST.

PDF ICLRW OpenReview Semantic Scholar

Cite

Text

Dognin et al. "Improved Adversarial Image Captioning." ICLR 2019 Workshops: DeepGenStruct, 2019.

Markdown

[Dognin et al. "Improved Adversarial Image Captioning." ICLR 2019 Workshops: DeepGenStruct, 2019.](https://mlanthology.org/iclrw/2019/dognin2019iclrw-improved/)

BibTeX

@inproceedings{dognin2019iclrw-improved,
  title     = {{Improved Adversarial Image Captioning}},
  author    = {Dognin, Pierre and Melnyk, Igor and Mroueh, Youssef and Ross, Jarret and Sercu, Tom},
  booktitle = {ICLR 2019 Workshops: DeepGenStruct},
  year      = {2019},
  url       = {https://mlanthology.org/iclrw/2019/dognin2019iclrw-improved/}
}