Abstract
Recent work in open-domain conversational agents has demonstrated that significant improvements in humanness and user preference can be achieved via massive scaling in both pre-training data and model size (Adiwardana et al., 2020; Roller et al., 2020). However, if we want to build agents with human-like abilities, we must expand beyond handling just text. A particularly important topic is the ability to see images and communicate about what is perceived. With the goal of getting humans to engage in multi-modal dialogue, we investigate combining components from state-of-the-art open-domain dialogue agents with those from state-of-the-art vision models. We study incorporating different image fusion schemes and domain-adaptive pre-training and fine-tuning strategies, and show that our best resulting model outperforms strong existing models in multi-modal dialogue while simultaneously performing as well as its predecessor (text-only) BlenderBot (Roller et al., 2020) in text-based conversation. We additionally investigate and incorporate safety components in our final model, and show that such efforts do not diminish model performance with respect to human preference.- Anthology ID:
- 2021.emnlp-main.398
- Volume:
- Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing
- Month:
- November
- Year:
- 2021
- Address:
- Online and Punta Cana, Dominican Republic
- Venue:
- EMNLP
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 4863–4883
- Language:
- URL:
- https://aclanthology.org/2021.emnlp-main.398
- DOI:
- 10.18653/v1/2021.emnlp-main.398
- Cite (ACL):
- Kurt Shuster, Eric Michael Smith, Da Ju, and Jason Weston. 2021. Multi-Modal Open-Domain Dialogue. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 4863–4883, Online and Punta Cana, Dominican Republic. Association for Computational Linguistics.
- Cite (Informal):
- Multi-Modal Open-Domain Dialogue (Shuster et al., EMNLP 2021)
- PDF:
- https://preview.aclanthology.org/ingestion-script-update/2021.emnlp-main.398.pdf
- Data
- Blended Skill Talk, COCO Captions, ConvAI2, EmpatheticDialogues, Wizard of Wikipedia