Hallucination Augmented Contrastive Learning for Multimodal Large Language Model
Authors:
Top Cited Papers
- 16 June 2024
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 27026-27036
Abstract
Multi-modal large language models (MLLMs) have been shown to efficiently integrate natural language with visual information to handle multi-modal tasks. However, MLLMs still face a fundamental limitation of hallucinations, where they tend to generate erroneous or fabricated information. In this paper, we address hallucinations in MLLMs from a novel perspective of representation learning. We first analyzed the representation distribution of textual and visual tokens in MLLM, revealing two important findings: 1) there is a significant gap between textual and visual representations, indicating unsatisfactory cross-modal representation alignment; 2) representations of texts that contain and do not contain hallucinations are entangled, making it challenging to distinguish them. These two observations inspire us with a simple yet effective method to mitigate hallucinations. Specifically, we introduce contrastive learning into MLLMs and use text with hallucination as hard negative examples, naturally bringing representations of non-hallucinative text and visual samples closer while pushing way representations of non-hallucinating and hallucinative text. We evaluate our method quantitatively and qualitatively, showing its effectiveness in reducing hallucination occurrences and improving performance across multiple benchmarks. On the MMhal-Bench benchmark, our method obtains a 34.66% /29.5% improvement over the baseline MiniGPT-4/LLaVA. Our code is available on https://github.com/X-PLUG/mPLUG-HalOwl/tree/main/hacl.Keywords
All Related Versions
This publication has 14 references indexed in Scilit:
- COPA : Efficient Vision-Language Pre-training through Collaborative Object- and Patch-Text Alignment2023
- BUS : Efficient and Effective Vision-language Pre-training with Bottom-Up Patch Summarization2023
- Vision-Language Pre-Training with Triple Contrastive Learning2022
- Conceptual 12M: Pushing Web-Scale Image-Text Pre-Training To Recognize Long-Tail Visual Concepts2021
- End-to-End Object Detection with Transformers2020
- ZeRO: Memory optimizations Toward Training Trillion Parameter Models2020
- Momentum Contrast for Unsupervised Visual Representation Learning2020
- OCR-VQA: Visual Question Answering by Reading Text in Images2019
- Towards VQA Models That Can Read2019
- Microsoft COCO: Common Objects in Context2014