ALOHa: A New Measure for Hallucination in Captioning Models

Suzanne Petryk, David Chan, Anish Kachinthaya, Haodi Zou, John Canny, Joseph Gonzalez, Trevor Darrell


Abstract
Despite recent advances in multimodal pre-training for visual description, state-of-the-art models still produce captions containing errors, such as hallucinating objects not present in a scene. The existing prominent metric for object hallucination, CHAIR, is limited to a fixed set of MS COCO objects and synonyms. In this work, we propose a modernized open-vocabulary metric, ALOHa, which leverages large language models (LLMs) to measure object hallucinations. Specifically, we use an LLM to extract groundable objects from a candidate caption, measure their semantic similarity to reference objects from captions and object detections, and use Hungarian matching to produce a final hallucination score. We show that ALOHa correctly identifies 13.6% more hallucinated objects than CHAIR on HAT, a new gold-standard subset of MS COCO Captions annotated for hallucinations, and 30.8% more on nocaps, where objects extend beyond MS COCO categories.
Anthology ID:
2024.naacl-short.30
Volume:
Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 2: Short Papers)
Month:
June
Year:
2024
Address:
Mexico City, Mexico
Editors:
Kevin Duh, Helena Gomez, Steven Bethard
Venue:
NAACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
342–357
Language:
URL:
https://aclanthology.org/2024.naacl-short.30
DOI:
Bibkey:
Cite (ACL):
Suzanne Petryk, David Chan, Anish Kachinthaya, Haodi Zou, John Canny, Joseph Gonzalez, and Trevor Darrell. 2024. ALOHa: A New Measure for Hallucination in Captioning Models. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 2: Short Papers), pages 342–357, Mexico City, Mexico. Association for Computational Linguistics.
Cite (Informal):
ALOHa: A New Measure for Hallucination in Captioning Models (Petryk et al., NAACL 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.naacl-short.30.pdf