AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

LVLM-EHub: A Comprehensive Evaluation Benchmark for Large Vision-Language Models

Peng Xu, Wenqi Shao, Kaipeng Zhang, Peng Gao, Shuo Liu, Meng Lei, Fanqing Meng, Siyuan Huang, Yu Qiao, Ping Luo

TPAMI 2024 · Vision-language and multimodal › Multimodal LLMs · approach: VLM capability analysis · 282 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
use multi-turn reasoning evaluation instead of single-turn QA on SNLI-VEinstruction-tuned LVLMs vs BLIP2 · SNLI-VEaccuracy: BLIP2 32.00 vs LLaVA 57.80, InstructBLIP 59.00Improved
multi-turn reasoning evaluation on VCRinstruction-tuned LVLMs vs BLIP2 · VCRaccuracy: BLIP2 31.60 vs InstructBLIP 54.20Improved
massive VQA instruction tuning (InstructBLIP) vs BLIP2InstructBLIP · visual perception, knowledge, reasoningaverage score: 0.858 -> 0.928 perception; 0.927 -> 0.967 knowledge acquisitionImproved
add massive in-domain VQA tuning (InstructBLIP) on embodied planningInstructBLIP · VirtualHomeaverage score: 0.922 (LA-V2) vs 0.772 (InstructBLIP)Hurt
instruction tuning with moderate instruction data on POPE hallucinationLLaVA, LA-V2, mPLUG-Owl, Otter vs BLIP2 · MSCOCO-Random POPEaccuracy: 82.21 (BLIP2) -> 51.52 (LLaVA), 74.44 (LA-V2)Hurt
change the prompt for image captioningVPGTrans · NoCapsCIDEr: 19.66 to 48.13Mixed
add visual instruction tuning (vs BLIP2) for embodied planningLLaMA-Adapter-v2, LLaVA · VirtualHomeexecutability: 2.88 -> 4.08 (LA-V2), 3.82 (LLaVA)Improved
add image-text pair pretraining (204M pairs) to instruction tuningmPLUG-Owl vs Otter · VirtualHomespatial relation score: 3.10 -> 3.22Improved
larger visual encoder and Q-Former trained on image-text pairsBLIP2, InstructBLIP, VPGTrans · OCR/KIE/captioningaverage score: 0.927, 0.967, 0.720 vs LLaVA 0.377Improved
update LLM with 158K instruction-following datamPLUG-Owl, LLaVA · ImageNet1K, CIFAR10accuracy: 26.68 (mPLUG-Owl ImageNet1K), 67.96 (LLaVA CIFAR10)Improved

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