1 of 12

Relevance-aware Multi-context Contrastive Decoding for Retrieval-augmented Visual Question Answering

Presented at WACV 2026�

Jongha Kim, Byungoh Ko, Jeehye Na, Jinsung Yoon, Hyunwoo J. Kim

1

Korea University

MLV Lab

NeurIPS 2024

2 of 12

Retrieval-Augmented Generation for Visual Question Answering

In Retrieval-Augmented Generation (RAG), relevant contexts are retrieved from external knowledge bases and fed to an MLLM to supplement missing or incomplete knowledge.

[1] Chen et al. ”Can Pre-trained Vision and Language Models Answer Visual Information-Seeking Questions”, EMNLP 2023

Question samples of preivous VQA dataset (top) and InfoSeek (bottom)

MLV Lab

Korea University

MLV Lab

NeurIPS 2024

3 of 12

Not all retrieved contexts are equally relevant

Although multiple contexts are retrieved, their relevance varies across samples: some are highly relevant, while others are only weakly relevant or entirely irrelevant.

[1] Wang et al. ”REAR: A Relevance-Aware Retrieval-Augmented Framework for Open-Domain Question Answering”, EMNLP 2024

Example of retrieved results with different level of relevance

MLV Lab

Korea University

MLV Lab

NeurIPS 2024

4 of 12

Goal: Effectively leveraging multiple retrieved contexts

Previous MLLMs treat every retrieved context equally, agnostic to their relevance. Thus, we propose a new decoding method leveraging multiple contexts considering their relevance.

Accuracy of baseline and baseline + RMCD (ours) by the number of retrieved contexts

MLV Lab

Korea University

MLV Lab

NeurIPS 2024

5 of 12

RMCD: Relevance-aware Multi-context Decoding

Using the retrieval scores, we compute context weights that reflect each context’s relative importance: relevant contexts receive larger weights, while irrelevant ones receive smaller or even negative weights.

Overall pipeline of RMCD

MLV Lab

Korea University

MLV Lab

NeurIPS 2024

6 of 12

RMCD: Relevance-aware Multi-context Decoding

Then, we aggregate the output logits from each retrieved context using the computed weights, amplifying the influence of relevant contexts while suppressing (or deflecting) irrelevant ones.

Overall pipeline of RMCD

MLV Lab

Korea University

MLV Lab

NeurIPS 2024

7 of 12

Quantitative Results

By replacing decoding methods of baselines, RMCD consistently improves accuracy across multiple MLLMs and knowledge-intensive benchmarks.

Accuracy on InfoSeek / Encyclopedic-VQA (left) and OK-VQA (right)

MLV Lab

Korea University

MLV Lab

NeurIPS 2024

8 of 12

Cost Analysis

While utilizing same number of contexts, RMCD maintain or improves latency, while improving QA accuracy. Moreover, RMCD does not require additional training, as it is only applied during inference.

Accuracy on OK-VQA (left) and latency comparison with baselines (right)

MLV Lab

Korea University

MLV Lab

NeurIPS 2024

9 of 12

Performance by retrieval quality

RMCD consistently outperforms every other decoding methods regardless of the retrieval quality, showing the best accuracy under both the best and worst retrieval results.

Retrieval results (left) and accuracy by decoding methods (right) by knowledge base sizes.�Retrieval gets difficult in larger knowledge base as portion of irrelevant corpus increases.

MLV Lab

Korea University

MLV Lab

NeurIPS 2024

10 of 12

Performance with oracle retriever

RMCD yields additional accuracy gains even with an oracle retriever, where the ground-truth is always ranked top-1, demonstrating that it complements retriever improvements.

Accuracy with oracle retriever, where ground-truth corpus is always set as top-1 retrieval result.

MLV Lab

Korea University

MLV Lab

NeurIPS 2024

11 of 12

Qualitative Results

Higher weights are assigned to relevant contexts, while negative weights are assigned to irrelevant contexts which distract baseline decoding methods.

MLV Lab

Korea University

MLV Lab

NeurIPS 2024

12 of 12

Conclusion

  • Previous MLLM treats all retrieved contexts equally, not considering each context’s relevance. Therefore, relevant contexts are underutilized, while irrelevant contexts negatively impacts results.�
  • We propose Relevance-aware Multi-context Contrastive Decoding (RMCD), which contrasts multiple contexts based on their relevance.

  • RMCD improves accuracy across multiple MLLMs and benchmarks without training by reflecting/deflecting relevant/irrelevant contexts during decoding phase.

Github

Paper

MLV Lab

Korea University

MLV Lab

NeurIPS 2024