Home News Latest Papers Tutorials Datasets Wiki SOTA LLM Models GPU Leaderboard Events

English

Visual Dialog On Visdial V09 Val

Metrics

MRR

Mean Rank

R@1

R@10

R@5

Results

Performance results of various models on this benchmark

Model Name	MRR	Mean Rank	R@1	R@10	R@5	Paper Title	Repository
CoAtt	63.98	4.47	50.29	88.81	80.71	Are You Talking to Me? Reasoned Visual Dialog Generation through Adversarial Learning	-
HACAN	0.6792	3.97	54.76	90.68	83.03	Making History Matter: History-Advantage Sequence Training for Visual Dialog	-
HRE-QIH-D	0.5846	5.72	44.67	84.22	74.50	Visual Dialog
DAN	66.38	4.04	53.33	90.38	82.42	Dual Attention Networks for Visual Reference Resolution in Visual Dialog
CorefNMN	63.6	4.53	50.24	88.51	79.81	Visual Coreference Resolution in Visual Dialog using Neural Module Networks
CAG	0.6756	3.75	54.64	91.48	83.72	Iterative Context-Aware Graph Inference for Visual Dialog
DualVD	62.94	4.17	48.64	89.94	80.89	DualVD: An Adaptive Dual Encoding Model for Deep Visual Understanding in Visual Dialogue
HRE-QIH-D	0.5807	5.78	43.82	84.07	74.68	Visual Dialog
MVAN	0.6765	3.73	54.65	91.47	83.85	Multi-View Attention Network for Visual Dialog
AMEM	-	4.86	48.53	87.43	78.66	Visual Reference Resolution using Attention Memory for Visual Dialog	-
CorefNMN (ResNet-152)	64.1	4.45	50.92	88.81	80.18	Visual Coreference Resolution in Visual Dialog using Neural Module Networks
RVA	0.6634	3.93	52.71	90.73	82.97	Recursive Visual Attention in Visual Dialog
HCIAE-NP-ATT	62.22	4.81	48.48	87.59	78.75	Best of Both Worlds: Transferring Knowledge from Discriminative Learning to a Generative Visual Dialog Model
9xFGA (VGG)	68.92	3.39	55.16	92.95	86.26	Factor Graph Attention
HieCoAtt-QI	57.88	5.84	43.51	83.96	74.49	Hierarchical Question-Image Co-Attention for Visual Question Answering
MN-QIH-D	0.5965	5.46	45.55	85.37	76.22	Visual Dialog
GNN	0.6285	4.57	48.95	88.36	79.65	Reasoning Visual Dialogs with Structural and Partial Observations
SF-QIH-se-2	62.42	4.70	48.55	87.75	78.96	Two can play this Game: Visual Dialog with Discriminative Question Generation and Answering	-

0 of 18 row(s) selected.