List your company
ResearchFormerly WizWand

Image Quality Assessment on AGIQA-3K

0.9091SRCCMST-CLIPIQA
0.6760360.7365430.797050.857557Mar 17, 2024Aug 4, 2024Dec 22, 2024May 11, 2025Sep 28, 2025Feb 15, 2026Jul 6, 2026
Updated 3mo ago

Evaluation results

MethodLinks
MST-CLIPIQA2026.060.90910.9282
MST-CLIPIQA*2026.060.90850.9283
LIQE2026.060.90090.922
MA-AGIQA2026.060.89390.9273
MA-AGIQAType=data-driven, Supervision Setting=Supervised2026.020.8930.927
MANIQA2026.060.89160.9194
AGQG +IPCEEvaluation Protocol=Intra-dataset, Incorporation=AGQG module2025.120.89110.9284
PKT-12026.070.89080.9205
PKT-22026.070.88870.9194
LIQE2026.070.88790.9174
AGQG +CLIP-AGIQAEvaluation Protocol=Intra-dataset, Incorporation=AGQG module2025.120.88640.9268
IPCEEvaluation Protocol=Intra-dataset2025.120.88410.9246
CLIP-AGIQA2026.070.87930.9127
ELIQType=data-driven, Supervision Setting=Weak-supervised2026.020.8760.911
ViLEvaluation Protocol=Intra-dataset2025.120.8750.9145
ViT/B/16Evaluation Protocol=Intra-dataset2025.120.86590.9115
CLIP-AGIQAEvaluation Protocol=Intra-dataset2025.120.86180.8978
AGQG +CLIP-IQAEvaluation Protocol=Intra-dataset, Incorporation=AGQG module2025.120.86140.9106
MANIQAType=data-driven, Supervision Setting=Supervised2026.020.8610.911
QualiCLIP+Opinion-Unaware=false2024.030.860.903
MANIQA2026.070.85660.901
AMFF-NetEvaluation Protocol=Intra-dataset2025.120.85650.905
ResNet50Evaluation Protocol=Intra-dataset2025.120.85520.9072
StairIQA2026.060.85430.8943
HyperIQAEvaluation Protocol=Intra-dataset2025.120.85260.8975
SSL-IPFEvaluation Protocol=Intra-dataset2025.120.85230.9045
Q-AlignType=data-driven, Supervision Setting=Supervised2026.020.8520.881
HyperIQAType=data-driven, Supervision Setting=Supervised2026.020.850.904
HyperIQA2026.060.84950.8923
StairIQA2026.070.84710.8954
CLIP-IQA+Opinion-Unaware=false2024.030.8440.894
CLIP-IQA+Type=data-driven, Supervision Setting=Weak-supervised2026.020.8440.894
StairIQAEvaluation Protocol=Intra-dataset2025.120.84390.8989
HyperIQA2026.070.84320.8924
CLIP-IQAEvaluation Protocol=Intra-dataset2025.120.84260.8053
AGQG +TIEREvaluation Protocol=Intra-dataset, Incorporation=AGQG module2025.120.84210.8977
StairIQAType=data-driven, Supervision Setting=Supervised2026.020.8340.893
MUSIQ2026.060.83380.8698
MGQAEvaluation Protocol=Intra-dataset2025.120.82830.8944
DBCNNType=data-driven, Supervision Setting=Supervised2026.020.8260.89
TIEREvaluation Protocol=Intra-dataset2025.120.82510.8821
MG-IQATraining Dataset=KADID-10k, Evaluation=Zero-shot2026.040.8240.841
MUSIQType=data-driven, Supervision Setting=Supervised2026.020.820.865
LinearityIQA2026.060.81890.8309
MUSIQ2026.070.81880.8692
CONTRIQUEOpinion-Unaware=false2024.030.8170.879
CONTRIQUEType=data-driven, Supervision Setting=Weak-supervised2026.020.8170.879
VGG16Evaluation Protocol=Intra-dataset2025.120.81670.8752
DBCNNEvaluation Protocol=Intra-dataset2025.120.81540.8747
MST-CLIPIQA*prompts=with2026.060.81240.8817
Re-IQAOpinion-Unaware=false2024.030.8110.874
Re-IQAType=data-driven, Supervision Setting=Weak-supervised2026.020.8110.874
GRepQOpinion-Unaware=false2024.030.8070.858
GRepQ-DType=data-driven, Supervision Setting=Weak-supervised2026.020.8070.858
LinearityIQA2026.070.80430.8191
ARNIQAOpinion-Unaware=false2024.030.8030.881
ARNIQAType=data-driven, Supervision Setting=Weak-supervised2026.020.8030.881
ELIQType=data-driven, Supervision Setting=Label-free2026.020.8010.827
VQ-R1Training Dataset=KADID-10k, Evaluation=Zero-shot2026.040.7970.814
HyperAlignVenue=-2026.010.79270.883
MST-CLIPIQAprompts=without2026.060.78950.8666
CIA-NetVenue=PR’252026.010.77970.8687
IPCEVenue=CVPRW’242026.010.76970.8725
Q-DeepSight2026.040.7680.822
Q-insightTraining Dataset=KonIQ, KADIS2026.010.7660.816
Q-insightVenue=NeurIPS’25, Methodology=MLLM-Based Methods2026.010.7660.816
Q-InsightTraining Dataset=KADID-10k, Evaluation=Zero-shot2026.040.7660.783
Q-Insight (regression)Method Category=MLLM-based: RL training, Backbone=Qwen2.5-VL-7B, Reproduction Protocol=Reported2026.060.7660.816
LIQE2026.060.76380.848
VisualQuality-R1Venue=NeurIPS’25, Methodology=MLLM-Based Methods2026.010.760.817
IP-IQAVenue=ICME’242026.010.75780.8544
VisualQuality-R12026.040.7540.82
Q-HawkeyeTraining Dataset=KonIQ2026.010.7520.807
Q-HawkeyeMethodology=MLLM-Based Methods2026.010.7520.807
AMFF-NetVenue=TBC’242026.010.75130.8476
AMFF-NetOriginal Results=Yes2026.060.75130.8476
Q-Insight2026.040.7490.81
DeQA-ScoreTraining Dataset=KonIQ, SPAQ, KADID2026.010.7450.808
DeQA-ScoreTraining Dataset=KADID-10k, Evaluation=Zero-shot2026.040.740.758
Qwen-SFTVenue=arXiv’25, Methodology=MLLM-Based Methods2026.010.7390.813
DeQA-ScoreTraining Dataset=KonIQ, SPAQ2026.010.7380.81
DeQA-Score2026.040.7380.79
CLIP-IQA+Venue=AAAI’23, Methodology=Non-MLLM Deep-Learning2026.010.7370.739
DeQA-ScoreTraining Dataset=KonIQ, SPAQ, KADID, PIPAL2026.010.7350.77
Q-AlignVenue=ICML’24, Methodology=MLLM-Based Methods2026.010.7350.772
Q-AlignMethod Category=MLLM-based: SFT training2026.060.7350.772
VisualQuality-R1Training Dataset=KADID, SPAQ2026.010.7330.8
MR-IQAMethod Category=MLLM-based: RL training, Backbone=Qwen3-VL-2B, Reproduction Protocol=Controlled reproduction2026.060.7320.804
Q-AlignTraining Dataset=KADID-10k, Evaluation=Zero-shot2026.040.7310.752
ImageReward2026.060.72970.7847
DeQA-ScoreVenue=CVPR’25, Methodology=MLLM-Based Methods2026.010.7290.809
DeQAMethod Category=MLLM-based: SFT training2026.060.7290.809
LIQETraining Dataset=KADID-10k, Evaluation=Zero-shot2026.040.7190.741
VQ-R1 (ranking)Method Category=MLLM-based: RL training, Backbone=Qwen3-VL-2B, Reproduction Protocol=Controlled reproduction2026.060.7180.744
TOPIQTraining Dataset=KADID-10k, Evaluation=Zero-shot2026.040.7130.731
MANIQATraining Dataset=KADID-10k, Evaluation=Zero-shot2026.040.7070.728
DBCNNVenue=ICSIPA’19, Methodology=Non-MLLM Deep-Learning2026.010.7010.73
Re-IQATraining Dataset=KADID-10k, Evaluation=Zero-shot2026.040.70.719
MUSIQTraining Dataset=KADID-10k, Evaluation=Zero-shot2026.040.6940.715
CLIP-IQA+Method Category=Deep-learning-based2026.060.6850.736