Visual Reasoning on NLVR2 Test

14개 결과 · ⬇ CSV · JSON

Accuracy

Rank	Model	Accuracy	Paper	Code	Year
1	BEiT-3	92.58	Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks	microsoft/unilm · lyan62/data-curation	2022
2	X2-VLM (large)	89.4	X$^2$-VLM: All-In-One Pre-trained Model For Vision-Language Tasks	zengyan-97/x-vlm · zengyan-97/x2-vlm	2022
3	XFM (base)	88.4	Toward Building General Foundation Models for Language, Vision, and Vision-Language Understanding Tasks	zhangxinsong-nlp/XFM	2023
4	CoCa	87.0	CoCa: Contrastive Captioners are Image-Text Foundation Models	mlfoundations/open_clip · facebookresearch/multimodal · lucidrains/CoCa-pytorch · +3	2022
4	X2-VLM (base)	87.0	X$^2$-VLM: All-In-One Pre-trained Model For Vision-Language Tasks	zengyan-97/x-vlm · zengyan-97/x2-vlm	2022
6	VLMo	86.86	VLMo: Unified Vision-Language Pre-Training with Mixture-of-Modality-Experts	microsoft/unilm · ylsung/vl-merging	2021
7	SimVLM	85.15	SimVLM: Simple Visual Language Model Pretraining with Weak Supervision	yulong-XJTU/SimVLM · FerryHuang/SimVLM	2021
8	X-VLM (base)	84.76	Multi-Grained Vision Language Pre-Training: Aligning Texts with Visual Concepts	zengyan-97/x-vlm	2021
9	BLIP-129M	83.09	BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation	huggingface/transformers · salesforce/lavis · salesforce/blip · +6	2022
10	ALBEF (14M)	82.55	Align before Fuse: Vision and Language Representation Learning with Momentum Distillation	salesforce/lavis · salesforce/ALBEF · facebookresearch/multimodal · +3	2021
11	UNITER (Large)	79.5	UNITER: UNiversal Image-TExt Representation Learning	ChenRocks/UNITER · YIKUAN8/Transformers-VQA · necla-ml/SNLI-VE · +4	2019
12	SOHO	77.32	Seeing Out of tHe bOx: End-to-End Pre-training for Vision-Language Representation Learning	researchmm/soho · PasserBy4/mypretraining · PasserBy4/pretraining	2021
13	LXMERT	76.2	LXMERT: Learning Cross-Modality Encoder Representations from Transformers	huggingface/transformers · airsplay/lxmert · zhegan27/VILLA · +6	2019
14	ViLT-B/32	76.13	ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision	huggingface/transformers · dandelin/vilt · glamor-usc/climb · +3	2021