Developed by AIRG, AiCUS is an AI software for automated photography aesthetics assessment.
Based on the architecture of multi-column deep convolutional neural network, AiCUS has been trained with 700,000 human curated photographs across more than 10 categories, and formed its own aesthetic criteria for assessing a photograph.
Most importantly, AiCUS models are being constantly trained, with new datasets and new training strategies, and hence its performance will continuously improve.
Currently, AiCUS provides two platfortms for photograph assessment:
AiCUS Landscape – assessing landscape photographs only
AiCUS Universal – assess all other category photographs
The score given by AiCUS ranges from 0 to 1:
[0.9, 1.000[ – Exceptional
[0.7, 0.8999[ – Excellent
[0.5, 0.6999] – Good
[0.4, 0.4999] – Fair
[0.0, 0.3999] – Poor
You can submit your photos here, to see how AiCUS evaluates them. AiCUS accepts all common photo formats such as jpeg, png, etc.. For achieving the best result, you may keep your photo between 1500 pixels – 3000 pixels for the longest side.