Evaluating and Analyzing Relationship Hallucinations in Large Vision-Language Models

Mingrui Wu, Jiayi Ji, Oucheng Huang, Jiale Li, Yuhang Wu, Xiaoshuai Sun, Rongrong Ji
Proceedings of the 41st International Conference on Machine Learning, PMLR 235:53553-53570, 2024.

Abstract

The issue of hallucinations is a prevalent concern in existing Large Vision-Language Models (LVLMs). Previous efforts have primarily focused on investigating object hallucinations, which can be easily alleviated by introducing object detectors. However, these efforts neglect hallucinations in inter-object relationships, which is essential for visual comprehension. In this work, we introduce R-Bench, a novel benchmark for evaluating Vision Relationship Hallucination. R-Bench features image-level questions that focus on the existence of relationships and instance-level questions that assess local visual comprehension. We identify three types of relationship co-occurrences that lead to hallucinations: relationship-relationship, subject-relationship, and relationship-object. The visual instruction tuning dataset’s long-tail distribution significantly impacts LVLMs’ understanding of visual relationships. Additionally, our analysis reveals that current LVLMs tend to overlook visual content, overly rely on the common sense knowledge of Large Language Models (LLMs), and struggle with spatial relationship reasoning based on contextual information.

Cite this Paper

BibTeX
@InProceedings{pmlr-v235-wu24l, title = {Evaluating and Analyzing Relationship Hallucinations in Large Vision-Language Models}, author = {Wu, Mingrui and Ji, Jiayi and Huang, Oucheng and Li, Jiale and Wu, Yuhang and Sun, Xiaoshuai and Ji, Rongrong}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, pages = {53553--53570}, year = {2024}, editor = {Salakhutdinov, Ruslan and Kolter, Zico and Heller, Katherine and Weller, Adrian and Oliver, Nuria and Scarlett, Jonathan and Berkenkamp, Felix}, volume = {235}, series = {Proceedings of Machine Learning Research}, month = {21--27 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v235/main/assets/wu24l/wu24l.pdf}, url = {https://proceedings.mlr.press/v235/wu24l.html}, abstract = {The issue of hallucinations is a prevalent concern in existing Large Vision-Language Models (LVLMs). Previous efforts have primarily focused on investigating object hallucinations, which can be easily alleviated by introducing object detectors. However, these efforts neglect hallucinations in inter-object relationships, which is essential for visual comprehension. In this work, we introduce R-Bench, a novel benchmark for evaluating Vision Relationship Hallucination. R-Bench features image-level questions that focus on the existence of relationships and instance-level questions that assess local visual comprehension. We identify three types of relationship co-occurrences that lead to hallucinations: relationship-relationship, subject-relationship, and relationship-object. The visual instruction tuning dataset’s long-tail distribution significantly impacts LVLMs’ understanding of visual relationships. Additionally, our analysis reveals that current LVLMs tend to overlook visual content, overly rely on the common sense knowledge of Large Language Models (LLMs), and struggle with spatial relationship reasoning based on contextual information.} }
Endnote
%0 Conference Paper %T Evaluating and Analyzing Relationship Hallucinations in Large Vision-Language Models %A Mingrui Wu %A Jiayi Ji %A Oucheng Huang %A Jiale Li %A Yuhang Wu %A Xiaoshuai Sun %A Rongrong Ji %B Proceedings of the 41st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Ruslan Salakhutdinov %E Zico Kolter %E Katherine Heller %E Adrian Weller %E Nuria Oliver %E Jonathan Scarlett %E Felix Berkenkamp %F pmlr-v235-wu24l %I PMLR %P 53553--53570 %U https://proceedings.mlr.press/v235/wu24l.html %V 235 %X The issue of hallucinations is a prevalent concern in existing Large Vision-Language Models (LVLMs). Previous efforts have primarily focused on investigating object hallucinations, which can be easily alleviated by introducing object detectors. However, these efforts neglect hallucinations in inter-object relationships, which is essential for visual comprehension. In this work, we introduce R-Bench, a novel benchmark for evaluating Vision Relationship Hallucination. R-Bench features image-level questions that focus on the existence of relationships and instance-level questions that assess local visual comprehension. We identify three types of relationship co-occurrences that lead to hallucinations: relationship-relationship, subject-relationship, and relationship-object. The visual instruction tuning dataset’s long-tail distribution significantly impacts LVLMs’ understanding of visual relationships. Additionally, our analysis reveals that current LVLMs tend to overlook visual content, overly rely on the common sense knowledge of Large Language Models (LLMs), and struggle with spatial relationship reasoning based on contextual information.
APA
Wu, M., Ji, J., Huang, O., Li, J., Wu, Y., Sun, X. & Ji, R.. (2024). Evaluating and Analyzing Relationship Hallucinations in Large Vision-Language Models. Proceedings of the 41st International Conference on Machine Learning, in Proceedings of Machine Learning Research 235:53553-53570 Available from https://proceedings.mlr.press/v235/wu24l.html.

Related Material