AI Paper Atlas

Research map / Vision-language and multimodal / Visual question answering

Fine-grained Late-interaction Multi-modal Retrieval for Retrieval Augmented Visual Question Answering

Weizhe Lin, Jinghong Chen, Jingbiao Mei, Alexandru Coca, Bill Byrne

NeurIPS 2023 · Vision-language and multimodal › Visual question answering · approach: Universal VQA models

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
upgrade retriever from DPR to FLMR (late interaction)RA-VQA-v2 (T5-large) · OK-VQAVQA Score: 51.45 -> 54.85Improved
add retrieval augmentation to BLIP 2BLIP 2 · OK-VQAVQA Score: 55.44 -> 62.08Improved
remove late interaction (use DPR) (ablation)RA-VQA-v2 (BLIP 2) · OK-VQAVQA Score: 58.20 -> 62.08Improved
add region-of-interest features (ablation)RA-VQA-v2 (BLIP 2) · OK-VQAVQA Score: 60.41 -> 62.08Improved
add text-based vision (captions, OCR, objects)FLMR · OK-VQA GSPRRecall@5: 74.81 -> 85.99Improved
add feature-based vision (vision encoder)FLMR · OK-VQA GSPRRecall@5: 74.81 -> 85.08Improved
combine text-based and feature-based visionFLMR · OK-VQA GSPRRecall@5: 85.99 -> 87.02Improved
add 9 ROIs to the retrieverFLMR · OK-VQA GSPRRecall@5: 87.02 -> 89.32Improved
add 9 ROIs to the retrieverDPR · OK-VQA GSPRRecall@5: 83.43 -> 82.90Hurt
remove vision-language alignment pre-training (ablation)FLMR · OK-VQA GSPRRecall@5: 85.71 -> 89.32Improved
increase retrieved documents K from 5 to 50DPR + T5-large · OK-VQAVQA Score: 51.5 -> 52.1Mixed
increase retrieved documents K from 5 to 50FLMR + T5-large · OK-VQAVQA Score: 54.9 -> 55.4No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy