AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

LLMGA: Multimodal Large Language Model based Generation Assistant

bin xia, Shiyin Wang, Yingfan Tao, Yitong Wang, Jiaya Jia

ECCV 2024 · Vision-language and multimodal › Multimodal LLMs · approach: Unified any-modality language models · 19 citations (OpenAlex)

Read the paper

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
finetune SD1.5 in the second stage to follow detailed MLLM prompts (ablation)LLMGA-7b · MSCOCO T2IFID: 23.5946 -> 18.5234Improved
add second-stage SD finetuning (ablation)LLMGA-7b · Places outpainting (wide masks)FID: 3.2630 -> 2.4409Improved
use LLMGA-7B prompts instead of plain SD1.5SD1.5 · MSCOCO T2IFID: 24.0081 -> 18.5234Improved
scale MLLM from 7B to 13B (ablation)LLMGA · MSCOCO T2IFID: 18.5234 -> 18.4063Improved
add reference input (masked image) to restoration network (ablation)DiffIR on LLMGA outpainting · Places outpaintingFID: 2.3587 -> 2.2993Improved
add random brightness/contrast degradation in restoration training (ablation)DiffRIR · Places outpaintingFID: 2.2993 -> 2.2687Improved
control SD with detailed language prompt instead of fixed-size embedding (ablation)LLMGA · MMBenchaccuracy: 35.3 -> 62.9Improved
control SD with language prompt instead of embedding over multiple conversation turnsLLMGA · MSCOCO multi-turn generationimage qualityImproved
train with full prompt refinement data instead of 10% (ablation)LLMGA · MSCOCO T2IFID: 21.4460 -> 18.5234Improved
train with full inpainting & outpainting data instead of 10% (ablation)LLMGA · Places outpaintingFID: 3.1054 -> 2.4409Improved
train with full similar image generation data instead of 10% (ablation)LLMGA · MSCOCO T2IFID: 19.5698 -> 18.5234Improved
use LLMGA zero-shot for editing instead of training SD on editing data (vs MagicBrush)LLMGA · MagicBrush testL1: 0.0647 (MagicBrush) vs 0.0814 (LLMGA)Mixed

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