Visual Reasoning On Nlvr2 Dev

Métriques

Accuracy

Résultats

Résultats de performance de divers modèles sur ce benchmark

		Paper Title
BEiT-3	91.51	Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks
X2-VLM (large)	88.7	X$^2$-VLM: All-In-One Pre-trained Model For Vision-Language Tasks
XFM (base)	87.6	Toward Building General Foundation Models for Language, Vision, and Vision-Language Understanding Tasks
X2-VLM (base)	86.2	X$^2$-VLM: All-In-One Pre-trained Model For Vision-Language Tasks
CoCa	86.1	CoCa: Contrastive Captioners are Image-Text Foundation Models
VLMo	85.64	VLMo: Unified Vision-Language Pre-Training with Mixture-of-Modality-Experts
VK-OOD	84.6	Implicit Differentiable Outlier Detection Enable Robust Deep Multimodal Analysis
SimVLM	84.53	SimVLM: Simple Visual Language Model Pretraining with Weak Supervision
X-VLM (base)	84.41	Multi-Grained Vision Language Pre-Training: Aligning Texts with Visual Concepts
VK-OOD	83.9	Differentiable Outlier Detection Enable Robust Deep Multimodal Analysis
ALBEF (14M)	83.14	Align before Fuse: Vision and Language Representation Learning with Momentum Distillation
SOHO	76.37	Seeing Out of tHe bOx: End-to-End Pre-training for Vision-Language Representation Learning
ViLT-B/32	75.7	ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision
LXMERT (Pre-train + scratch)	74.9	LXMERT: Learning Cross-Modality Encoder Representations from Transformers
VisualBERT	66.7	VisualBERT: A Simple and Performant Baseline for Vision and Language

0 of 15 row(s) selected.

Visual Reasoning On Nlvr2 Dev | SOTA | HyperAI