Is Your Text-to-Image Model Robust to Caption Noise?
Journal:
arXiv
Published Date:
Dec 27, 2024
Abstract
In text-to-image (T2I) generation, a prevalent training technique involves
utilizing Vision Language Models (VLMs) for image re-captioning. Even though
VLMs are known to exhibit hallucination, generating descriptive content that
deviates from the visual reality, the ramifications of such caption
hallucinations on T2I generation performance remain under-explored. Through our
empirical investigation, we first establish a comprehensive dataset comprising
VLM-generated captions, and then systematically analyze how caption
hallucination influences generation outcomes. Our findings reveal that (1) the
disparities in caption quality persistently impact model outputs during
fine-tuning. (2) VLMs confidence scores serve as reliable indicators for
detecting and characterizing noise-related patterns in the data distribution.
(3) even subtle variations in caption fidelity have significant effects on the
quality of learned representations. These findings collectively emphasize the
profound impact of caption quality on model performance and highlight the need
for more sophisticated robust training algorithm in T2I. In response to these
observations, we propose a approach leveraging VLM confidence score to mitigate
caption noise, thereby enhancing the robustness of T2I models against
hallucination in caption.
