AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Groma: Localized Visual Tokenization for Grounding Multimodal Large Language Models

Chuofan Ma, Yi Jiang, Jiannan Wu, Zehuan Yuan, Xiaojuan Qi

ECCV 2024 · Vision-language and multimodal › Multimodal LLMs · approach: Instruction-tuned multimodal assistants · 88 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
replace CLIP with DINOv2 as image encoder for localization (ablation)DDETR head on frozen backbone · COCO detectionAP: 32.4 -> 38.9 (336x336)Improved
increase input resolution from 336 to 448 for DINOv2 (ablation)DDETR head on DINOv2 · COCO detectionAP: 38.9 -> 43.6Improved
unfreeze the LLM during instruction finetuning (ablation)Groma · Visual Genome region captioningCIDEr: 148.0 -> 158.4Improved
unfreeze the LLM during instruction finetuning (ablation)Groma · REC benchmarksaverage accuracy: 84.02 -> 86.52Improved
merge every four neighboring image tokens before the LLM (ablation)Groma · REC benchmarksaverage accuracy: 86.55 -> 86.47No clear effect
localized visual tokenization instead of LLM coordinate output or external modulesGroma · LVIS-GroundAR: 16.8 -> 28.8 vs FerretImproved
use DINOv2 image tokens instead of CLIP-aligned tokens in the MLLMGroma · LLaVA-Bench (COCO)conversation and reasoning score: Conversation 82.6 vs LLaVA 85.4; Reasoning 88.8 vs 92.1Hurt
frozen-LLM Groma versus prior region models (ablation)Groma (frozen LLM) · Visual Genome region captioningCIDEr: 148.0 vs GPT4RoI 145.2Improved

Later papers that compared with this method

6 later accepted papers report Groma-7B as a baseline in their results tables; 6 of them report a better number on most of the results they share with it.

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 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