Aesthetic Assessment on AVA (test)
0.865SRCCQPT V2
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| QPT V2Finetuning data percentage=100%2024.07 | 0.865 | — | — | — | — | — | — | — | — | 0.875 | — | |
| Q-ALIGN#Training=236K (92%), Extra Data?=false2023.12 | 0.822 | — | — | — | — | — | — | — | — | 0.817 | — | |
| Q-Align2024.07 | 0.822 | — | — | — | — | — | — | — | — | 0.817 | — | |
| LIQE#Training=236K (92%), Extra Data?=false2023.12 | 0.776 | — | — | — | — | — | — | — | — | 0.763 | — | |
| FEWSHOT-Q-ALIGN#Training=26K (10%), Extra Data?=false2023.12 | 0.776 | — | — | — | — | — | — | — | — | 0.775 | — | |
| LIQE2024.07 | 0.776 | — | — | — | — | — | — | — | — | 0.763 | — | |
| VILA-REvaluation Protocol=Supervised (Finetuned Rank-based adapter)2023.03 | 0.774 | — | — | — | — | — | — | — | — | 0.774 | — | |
| VILA#Training=236K (92%), Extra Data?=true2023.12 | 0.774 | — | — | — | — | — | — | — | — | 0.774 | — | |
| VILA2024.07 | 0.774 | — | — | — | — | — | — | — | — | 0.774 | — | |
| QPT V2Finetuning data percentage=60%2024.07 | 0.766 | — | — | — | — | — | — | — | — | 0.78 | — | |
| GATx3-GATPEvaluation Protocol=Supervised2023.03 | 0.762 | — | — | — | — | — | — | — | — | 0.764 | — | |
| GAT×3-GATP2024.07 | 0.762 | — | — | — | — | — | — | — | — | 0.764 | — | |
| TANetEvaluation Protocol=Supervised2023.03 | 0.758 | — | — | — | — | — | — | — | — | 0.765 | — | |
| TANet2024.07 | 0.758 | — | — | — | — | — | — | — | — | 0.765 | — | |
| Pool-3FCNetwork Configuration=Pool-3FC2019.04 | 0.756 | 81.72 | — | — | — | — | — | — | — | 0.757 | — | |
| MLSP (Pool-3FC)Evaluation Protocol=Supervised2023.03 | 0.756 | — | — | — | — | — | — | — | — | 0.757 | — | |
| MLSP#Training=236K (92%), Extra Data?=false2023.12 | 0.756 | — | — | — | — | — | — | — | — | 0.757 | — | |
| Hosu et al.Number of crops=202021.08 | 0.756 | 81.7 | — | — | — | — | — | — | — | 0.757 | — | |
| MLSP2024.07 | 0.756 | — | — | — | — | — | — | — | — | 0.757 | — | |
| Niu et al.Evaluation Protocol=Supervised2023.03 | 0.734 | — | — | — | — | — | — | — | — | 0.74 | — | |
| Hentschel et al.Evaluation Protocol=Supervised2023.03 | 0.731 | — | — | — | — | — | — | — | — | 0.741 | — | |
| MUSIQEvaluation Protocol=Supervised2023.03 | 0.726 | — | — | — | — | — | — | — | — | 0.738 | — | |
| MUSIQ#Training=236K (92%), Extra Data?=false2023.12 | 0.726 | — | — | — | — | — | — | — | — | 0.738 | — | |
| MUSIQEvaluation mode=multi-scale ensemble2021.08 | 0.726 | 81.5 | — | — | — | — | — | — | — | 0.738 | 0.242 | |
| MUSIQ2024.07 | 0.726 | — | — | — | — | — | — | — | — | 0.738 | — | |
| Aesthetic Predictor#Training=236K (92%), Extra Data?=false2023.12 | 0.721 | — | — | — | — | — | — | — | — | 0.723 | — | |
| Aesthetic Predictor2024.07 | 0.721 | — | — | — | — | — | — | — | — | 0.723 | — | |
| Zeng et al. (resnet101)Evaluation Protocol=Supervised2023.03 | 0.719 | — | — | — | — | — | — | — | — | 0.72 | — | |
| Zeng et al.Backbone=ResNet1012021.08 | 0.719 | 80.8 | — | — | — | — | — | — | — | 0.72 | 0.275 | |
| MUSIQ-singleEvaluation mode=single2021.08 | 0.719 | 81.4 | — | — | — | — | — | — | — | 0.731 | 0.247 | |
| AMPEvaluation Protocol=Supervised2023.03 | 0.709 | — | — | — | — | — | — | — | — | — | — | |
| AMP2021.08 | 0.709 | 80.3 | — | — | — | — | — | — | — | — | 0.279 | |
| MaxViTEvaluation Protocol=Supervised2023.03 | 0.708 | — | — | — | — | — | — | — | — | 0.745 | — | |
| MaxViT#Training=236K (92%), Extra Data?=false2023.12 | 0.708 | — | — | — | — | — | — | — | — | 0.745 | — | |
| MaxViT2024.07 | 0.708 | — | — | — | — | — | — | — | — | 0.745 | — | |
| VILA-PEvaluation Protocol=Zero-shot Learning, Prompting Strategy=ensemble prompts2023.03 | 0.657 | — | — | — | — | — | — | — | — | 0.663 | — | |
| AFDC + SPPEvaluation Protocol=Supervised2023.03 | 0.649 | — | — | — | — | — | — | — | — | 0.671 | — | |
| AFDC + SPPWarping strategy=4 warps2021.08 | 0.649 | 83.2 | — | — | — | — | — | — | — | 0.671 | 0.271 | |
| AFDC2024.07 | 0.649 | — | — | — | — | — | — | — | — | 0.671 | — | |
| AFDC + SPPWarping strategy=single warp2021.08 | 0.648 | 83 | — | — | — | — | — | — | — | — | 0.273 | |
| Talebi et al. [22]Model Version=V3, Retrained=true2019.04 | 0.639 | 72.3 | — | — | — | — | — | — | — | 0.645 | — | |
| CLIP-IQA+#Training=236K (92%), Extra Data?=false2023.12 | 0.619 | — | — | — | — | — | — | — | — | 0.586 | — | |
| CLIP-IQA+2024.07 | 0.619 | — | — | — | — | — | — | — | — | 0.586 | — | |
| Talebi et al. [22]Model Version=V22019.04 | 0.612 | 81.51 | — | — | — | — | — | — | — | 0.636 | — | |
| NIMA (Inception-v2)Evaluation Protocol=Supervised2023.03 | 0.612 | — | — | — | — | — | — | — | — | 0.636 | — | |
| NIMA#Training=236K (92%), Extra Data?=false2023.12 | 0.612 | — | — | — | — | — | — | — | — | 0.636 | — | |
| NIMABackbone=Inception-v22021.08 | 0.612 | 81.5 | — | — | — | — | — | — | — | 0.636 | — | |
| NIMA2024.07 | 0.612 | — | — | — | — | — | — | — | — | 0.636 | — | |
| VILA-PEvaluation Protocol=Zero-shot Learning, Prompting Strategy=single prompt2023.03 | 0.605 | — | — | — | — | — | — | — | — | 0.617 | — | |
| NIMABackbone=VGG162021.08 | 0.592 | 80.6 | — | — | — | — | — | — | — | 0.61 | — | |
| Kong et al.2019.04 | 0.558 | 77.33 | — | — | — | — | — | — | — | — | — | |
| Kong et al.Evaluation Protocol=Supervised2023.03 | 0.558 | — | — | — | — | — | — | — | — | — | — | |
| Kong et al.2021.08 | 0.558 | 77.3 | — | — | — | — | — | — | — | — | — | |
| A-Lamp2017.04 | — | 82.5 | 0.92 | — | — | — | — | — | — | — | — | |
| A-LampNumber of crops=502021.08 | — | 82.5 | — | — | — | — | — | — | — | — | — | |
| AVA2017.04 | — | 67 | — | — | — | — | — | — | — | — | — | |
| Brain Inspired Deep Neural Network (BDN)Network=BDN2019.11 | — | 78.08 | — | — | — | — | — | — | — | — | — | |
| DCNN2017.04 | — | 73.25 | — | — | — | — | — | — | — | — | — | |
| DMA-Net-ImgFu2017.04 | — | 75.4 | — | — | — | — | — | — | — | — | — | |
| Double Column Network (DCNN)Network=DCNN2019.11 | — | 73.25 | — | — | — | — | — | — | — | — | — | |
| Hii et al.2019.04 | — | 75.76 | — | — | — | — | — | — | — | — | — | |
| Kao et al.2017.09 | — | — | — | 71.42 | — | — | — | — | — | — | — | |
| Kao et al.2019.04 | — | 71.42 | — | — | — | — | — | — | — | — | — | |
| Kao et al. (2016)2017.09 | — | — | — | 76.58 | — | — | — | — | — | — | — | |
| Kong et al.2017.09 | — | — | — | 77.33 | — | 0.558 | — | — | — | — | — | |
| Lu et al. (2014)2017.09 | — | — | — | 74.46 | — | — | — | — | — | — | — | |
| Lu et al. (2015)2017.09 | — | — | — | 75.42 | — | — | — | — | — | — | — | |
| Lu et al. [11]2019.04 | — | 75.42 | — | — | — | — | — | — | — | — | — | |
| Lu et al. [12]2019.04 | — | 74.46 | — | — | — | — | — | — | — | — | — | |
| Ma et al.2017.09 | — | — | — | 81.7 | — | — | — | — | — | — | — | |
| Ma et al.2019.04 | — | 81.7 | — | — | — | — | — | — | — | — | — | |
| Mai et al.2017.09 | — | — | — | 77.1 | — | — | — | — | — | — | — | |
| Mai et al.2019.04 | — | 77.4 | — | — | — | — | — | — | — | — | — | |
| MNA-CNN2017.04 | — | 77.1 | 0.85 | — | — | — | — | — | — | — | — | |
| MNA-CNN-SceneVariant=Scene2017.04 | — | 77.4 | — | — | — | — | — | — | — | — | — | |
| MNA-CNN-Scene2021.08 | — | 76.5 | — | — | — | — | — | — | — | — | — | |
| MPadaNumber of crops=≥322021.08 | — | 83 | — | — | — | — | — | — | — | — | — | |
| Murray et al.2017.09 | — | — | — | 66.7 | — | — | — | — | — | — | — | |
| Murray et al.2019.04 | — | 66.7 | — | — | — | — | — | — | — | — | — | |
| New-MP-NetSubnet=Multi-Patch subnet, Adaptive Patch Selection=True2017.04 | — | 81.7 | 0.91 | — | — | — | — | — | — | — | — | |
| NIMABackbone=MobileNet2017.09 | — | — | — | 80.36 | 0.518 | 0.51 | 0.152 | 0.137 | 0.081 | — | — | |
| NIMABackbone=VGG162017.09 | — | — | — | 80.6 | 0.61 | 0.592 | 0.205 | 0.202 | 0.052 | — | — | |
| NIMABackbone=Inception-v22017.09 | — | — | — | 81.51 | 0.636 | 0.612 | 0.233 | 0.218 | 0.05 | — | — | |
| Signle Column Network (SCNN)Network=SCNN2019.11 | — | 71.2 | — | — | — | — | — | — | — | — | — | |
| SPP-CNN2017.04 | — | 76 | 0.84 | — | — | — | — | — | — | — | — | |
| Triple Column NetworkNetwork=Triple Column Network2019.11 | — | 82.3 | — | — | — | — | — | — | — | — | — | |
| VGG-Center-CropBackbone=VGG, Cropping Strategy=Center-Crop2017.04 | — | 72.2 | 0.83 | — | — | — | — | — | — | — | — | |
| VGG-PadBackbone=VGG, Strategy=Pad2017.04 | — | 72.9 | 0.83 | — | — | — | — | — | — | — | — | |
| VGG-WrapBackbone=VGG, Strategy=Wrap2017.04 | — | 74.1 | 0.84 | — | — | — | — | — | — | — | — | |
| Wang et al.2017.09 | — | — | — | 76.8 | — | — | — | — | — | — | — |