Incremental Learning for Compressed Pornographic Image Recognition

Abstract

A compressed pornographic image recognition method is proposed by using incremental learning. For describing pornographic image, visual words are created from low-resolution (LR) image reconstructed from the compressed stream of the pornographic image. Covering algorithm is utilized to train and recognize the visual words in order to build the initial classification model of pornographic image. At last, incremental learning is adopted to continuously adjust the classification rules to recognize the new pornographic image samples. The experimental results show that the proposed incremental learning method for compressed pornographic image has higher recognition rate as well as costs less recognition time.