Photo Filter Classification and Filter Recommendation without Much Manual Labeling

Because how users employ filters to photos may reveal user's preference or mental state, a photo filter classification method is potentially demanded to enable future large-scale analysis. We adopt the transfer learning technique to transform deep models pre-trained for object classification into models suitable for photo filter classification. Based on accurate classification results, we build a filter recommendation approach without much manual labeling. It can be easily extended when more training data are available. A series of experimental studies are conducted to demonstrate effectiveness of filter classification with transfer learning. We also demonstrate the proposed filter recommendation achieves encouraging performance.

I. Introduction

As the flourishing growth of social media platforms, tremendous amounts of photos like selfies and pictures taken in a journey are uploaded and shared online. To stylize photos or make them more attractive, various filters are employed by users. The study in [1] shows filtered photos would receive more views and comments. Because of the impact of photo filtering, many social platforms and mobile apps have provided such functionalities.