pyiqa.archs.laion_aes_arch

LAION-Aesthetics Predictor

Introduced by: https://github.com/christophschuhmann/improved-aesthetic-predictor

Module Contents

pyiqa.archs.laion_aes_arch.default_model_urls

class pyiqa.archs.laion_aes_arch.MLP(input_size, xcol='emb', ycol='avg_rating')

Bases: torch.nn.Module

forward(x)

class pyiqa.archs.laion_aes_arch.LAIONAes(pretrained=True, pretrained_model_path=None)

Bases: torch.nn.Module

LAIONAes is a class that implements a neural network architecture for image quality assessment.

The architecture is based on the ViT-L/14 model from the OpenAI CLIP library, and uses an MLP to predict image quality scores.

Parameters:

None

Returns:

A tensor representing the predicted image quality scores.

forward(x)