Distilling Knowledge from Object Classification to Aesthetics Assessment

TL;DR

This paper introduces a method to improve image aesthetics assessment by distilling semantic knowledge from multiple object classification models, enabling the model to better relate diverse contents to aesthetic labels.

Contribution

The paper proposes a novel knowledge distillation approach from pre-trained object classifiers to enhance aesthetics assessment accuracy.

Findings

Achieved 4.8% SRCC improvement over baseline.

Up to 7.2% SRCC gain on specific categories.

Outperformed 10 previous IAA methods.

Abstract

In this work, we point out that the major dilemma of image aesthetics assessment (IAA) comes from the abstract nature of aesthetic labels. That is, a vast variety of distinct contents can correspond to the same aesthetic label. On the one hand, during inference, the IAA model is required to relate various distinct contents to the same aesthetic label. On the other hand, when training, it would be hard for the IAA model to learn to distinguish different contents merely with the supervision from aesthetic labels, since aesthetic labels are not directly related to any specific content. To deal with this dilemma, we propose to distill knowledge on semantic patterns for a vast variety of image contents from multiple pre-trained object classification (POC) models to an IAA model. Expecting the combination of multiple POC models can provide sufficient knowledge on various image contents, the IAA model can easier learn to relate various distinct contents to a limited number of aesthetic labels. By supervising an end-to-end single-backbone IAA model with the distilled knowledge, the performance of the IAA model is significantly improved by 4.8% in SRCC compared to the version trained only with ground-truth aesthetic labels. On specific categories of images, the SRCC improvement brought by the proposed method can achieve up to 7.2%. Peer comparison also shows that our method outperforms 10 previous IAA methods.