Mitigating Open-Vocabulary Caption Hallucinations

Anonymous

Abstract: While recent years have seen rapid progress in image-conditioned text generation, image captioning still suffers from the fundamental issue of hallucinations, namely, the generation of spurious details that cannot be inferred from the given image. Existing methods largely use closed-vocabulary object lists to mitigate or evaluate hallucinations in image captioning, ignoring most types of hallucinations that occur in practice. To this end, we propose a framework for addressing hallucinations in image captioning in the open-vocabulary setting, including quantifying their presence and optimizing to mitigate such hallucinations. Our OpenCHAIR benchmark leverages generative foundation models to evaluate open-vocabulary caption hallucinations, surpassing the popular CHAIR benchmark in both diversity and accuracy. To mitigate open-vocabulary hallucinations on the sequence level, we propose MOCHa, an approach harnessing advancements in reinforcement learning. Our multi-objective reward function explicitly targets the trade-off between fidelity and adequacy in generations without requiring any strong supervision. MOCHa improves a large variety of image captioning models, as captured by our OpenCHAIR benchmark as well as other existing metrics. We will release our code, benchmark and models.

Paper Type: long

Research Area: Multimodality and Language Grounding to Vision, Robotics and Beyond

Contribution Types: Model analysis & interpretability, Publicly available software and/or pre-trained models

Languages Studied: English