Phrase Grounding On Flickr30K Entities Test

Metriken

R@1

Ergebnisse

Leistungsergebnisse verschiedener Modelle zu diesem Benchmark

		Paper Title
GLIPv2	87.7	GLIPv2: Unifying Localization and Vision-Language Understanding
FIBER-B	87.4	Coarse-to-Fine Vision-Language Pre-training with Fusion in the Backbone
GLIP	87.1	Grounded Language-Image Pre-training
PEVL	84.4	PEVL: Position-enhanced Pre-training and Prompt Tuning for Vision-language Models
MDETR-ENB5	84.3	MDETR -- Modulated Detection for End-to-End Multi-Modal Understanding
DIGN	78.73	Disentangled Motif-aware Graph Learning for Phrase Grounding
LCMCG	76.74	Learning Cross-modal Context Graph for Visual Grounding
Soft-Label Chain CRF (SL-CCRF)	74.69	Phrase Grounding by Soft-Label Chain Conditional Random Field
DDPN (ResNet-101)	73.3	Rethinking Diversified and Discriminative Proposal Generation for Visual Grounding
VisualBERT	71.33	VisualBERT: A Simple and Performant Baseline for Vision and Language
BAN (Bottom-Up detector)	69.69	Bilinear Attention Networks
MCB	48.69	Multimodal Compact Bilinear Pooling for Visual Question Answering and Visual Grounding
GroundeR 100.0% annot.	48.38	Grounding of Textual Phrases in Images by Reconstruction
DSPE	43.89	Learning Deep Structure-Preserving Image-Text Embeddings
CCA - Fast RCNN	41.77	Flickr30k Entities: Collecting Region-to-Phrase Correspondences for Richer Image-to-Sentence Models
CCA - VGG19	30.83	Flickr30k Entities: Collecting Region-to-Phrase Correspondences for Richer Image-to-Sentence Models
SCRC	27.8	Natural Language Object Retrieval
CCA	25.30	Flickr30k Entities: Collecting Region-to-Phrase Correspondences for Richer Image-to-Sentence Models

0 of 18 row(s) selected.