AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

BI-MDRG: Bridging Image History in Multimodal Dialogue Response Generation

Hee Suk Yoon, Eunseop Yoon, Joshua Tian Jin Tee, Kang Zhang, Yu-Jung Heo, Du-Seong Chang, Chang D. Yoo

ECCV 2024 · Vision-language and multimodal › Multimodal LLMs · approach: Unified any-modality language models · 2 citations (OpenAlex)

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
add multimodal causal attention mask modulation (ablation)BI-MDRG · ImageChat (zero-shot)BLEU-1: 10.0 -> 10.9Improved
feed image features through cross-attention instead of text-only image history (BI-MDRG vs Divter_LLM)Divter_LLM · ImageChat (zero-shot)BLEU-1: 8.6 -> 10.9Improved
replace Divter_LLM with BI-MDRG on MMDialog textual image descriptionDivter_LLM · MMDialogTextual image description BLEU-1: 44.2 -> 52.2Improved
replace Divter_LLM with BI-MDRG on MMDialog image qualityDivter_LLM · MMDialogImage response IS: 21.0 -> 22.4Improved
BI-MDRG (4B) vs MiniGPT5 (9B) on text response BLEU-1BI-MDRG · MMDialogText response BLEU-1: 27.6 vs 29.1Hurt
BI-MDRG vs retrieval/classifier intent models on image intent F1BI-MDRG · PhotoChatIntent F1: 55.7 vs 58.9 (T5-3B)No clear effect
add Citation Module instead of LLMCite for citation tags (ablation)BI-MDRG · MDICDINOv2 image consistency: 0.34 -> 0.53Improved
add LLMCite citation tags over no citations (4B) (ablation)BI-MDRG 4B · MDICDINOv2 image consistency: 0.25 -> 0.34Improved
add citation tags with customized text-to-image (ablation)BI-MDRG · MMDialogIntent/TID/TR dialogue response metrics: TR B1 27.4 -> 27.6No clear effect
scale VLM size from 4B to 9B without citation framework (ablation)BI-MDRG · MDICDINOv2 image consistency: 0.25 -> 0.26No clear effect
scale VLM size from 4B to 9B (ablation)BI-MDRG · MMDialogText response BLEU-1: 27.4 -> 30.3Improved
replace LLMCite with trained citation predictionDivter_LLM + LLMCite · MDICCitation token prediction accuracy: 33.5 -> 84.0Improved

Later papers that compared with this method

1 later accepted paper reports BI-MDRG as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

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