Open Domain Dialogue Generation with Latent Images


  • Ze Yang Beihang University
  • Wei Wu Meituan
  • Huang Hu Microsoft
  • Can Xu Microsoft
  • Wei Wang China Resources Group
  • Zhoujun Li Beihang University


Conversational AI/Dialog Systems, Language Grounding & Multi-modal NLP, Generation


We consider grounding open domain dialogues with images. Existing work assumes that both an image and a textual context are available, but image-grounded dialogues by nature are more difficult to obtain than textual dialogues. Thus, we propose learning a response generation model with both image-grounded dialogues and textual dialogues by assuming that the visual scene information at the time of a conversation can be represented by an image, and trying to recover the latent images of the textual dialogues through text-to-image generation techniques. The likelihood of the two types of dialogues is then formulated by a response generator and an image reconstructor that are learned within a conditional variational auto-encoding framework. Empirical studies are conducted in both image-grounded conversation and text-based conversation. In the first scenario, image-grounded dialogues, especially under a low-resource setting, can be effectively augmented by textual dialogues with latent images; while in the second scenario, latent images can enrich the content of responses and at the same time keep them relevant to contexts.




How to Cite

Yang, Z., Wu, W., Hu, H., Xu, C., Wang, W., & Li, Z. (2021). Open Domain Dialogue Generation with Latent Images. Proceedings of the AAAI Conference on Artificial Intelligence, 35(16), 14239-14247. Retrieved from



AAAI Technical Track on Speech and Natural Language Processing III