Abstract
Large-scale vision-language pre-trained (VLP) models are prone to hallucinate non-existent visual objects when generating text based on visual information. In this paper, we systematically study the object hallucination problem from three aspects. First, we examine recent state-of-the-art VLP models, showing that they still hallucinate frequently and models achieving better scores on standard metrics (e.g., CIDEr) could be more unfaithful. Second, we investigate how different types of image encoding in VLP influence hallucination, including region-based, grid-based, and patch-based. Surprisingly, we find that patch-based features perform the best and smaller patch resolution yields a non-trivial reduction in object hallucination. Third, we decouple various VLP objectives and demonstrate that token-level image-text alignment and controlled generation are crucial to reducing hallucination. Based on that, we propose a simple yet effective VLP loss named ObjMLM to further mitigate object hallucination. Results show that it reduces object hallucination by up to 17.4% when tested on two benchmarks (COCO Caption for in-domain and NoCaps for out-of-domain evaluation).- Anthology ID:
- 2023.eacl-main.156
- Volume:
- Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics
- Month:
- May
- Year:
- 2023
- Address:
- Dubrovnik, Croatia
- Editors:
- Andreas Vlachos, Isabelle Augenstein
- Venue:
- EACL
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 2136–2148
- Language:
- URL:
- https://aclanthology.org/2023.eacl-main.156
- DOI:
- 10.18653/v1/2023.eacl-main.156
- Cite (ACL):
- Wenliang Dai, Zihan Liu, Ziwei Ji, Dan Su, and Pascale Fung. 2023. Plausible May Not Be Faithful: Probing Object Hallucination in Vision-Language Pre-training. In Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics, pages 2136–2148, Dubrovnik, Croatia. Association for Computational Linguistics.
- Cite (Informal):
- Plausible May Not Be Faithful: Probing Object Hallucination in Vision-Language Pre-training (Dai et al., EACL 2023)
- PDF:
- https://preview.aclanthology.org/add_acl24_videos/2023.eacl-main.156.pdf