Table 2 Implicit NR-IQA algorithm performance comparison
From: Explicit-implicit dual stream network for image quality assessment
Algorithm | Depth | Indies on LIVE | Training method | Year | |
|---|---|---|---|---|---|
| Â | Â | SROCC | PLCC | Â | Â |
CNNIQA | 5 | 0.956 | 0.953 | CNN | 2014 |
DeepBIQ | 3 | 0.889 | 0.908 | CNN | 2017 |
RankIQA | 16 | 0.981 | 0.982 | CNN | 2017 |
dipIQ | 5 | 0.960 | 0.969 | RankNet | 2017 |
BIECON | 6 | 0.961 | 0.962 | CNN | 2017 |
deepIQA | 12 | 0.972 | 0.960 | VGGnet | 2018 |
MEON | 13 | 0.974 | 0.972 | CNN | 2018 |
BLINDER | 37 | 0.966 | 0.959 | VGGnet | 2018 |
DIQA | 7 | 0.970 | 0.971 | CNN | 2019 |
DeepVR-IQA | 7 | 0.882 | 0.888 | ResNet-50 | 2020 |