Image aesthetics quality evaluation method based on prompt learning

The invention discloses an image aesthetics quality evaluation method based on prompt learning, and relates to the technical field of image processing, and the method comprises the steps: obtaining an image data set, and constructing an image aesthetics quality evaluation model; inputting image data in the image data set into an image encoder to obtain image embedding; constructing positive and negative meaning prompt words and inputting the positive and negative meaning prompt words into a text encoder to obtain positive and negative meaning text embedding; calculating the cosine similarity between the image embedding and the positive and negative text embedding; performing normalization processing on the cosine similarity by adopting a normalization exponential function to obtain an aesthetic quality score; according to the aesthetic quality score, using a loss function to optimize and adjust parameters of the image aesthetic quality evaluation model, and obtaining a trained image aesthetic quality evaluati