Image Annotation by Object Hypotheses-oriented Deep Neural Networks

Image annotation generates a set of semantic labels that describe the contents of an input image. Recently deep learning techniques have achieved significant success in many areas of image processing. In this paper, we present a multi-label image annotation method that combines unsupervised object hypotheses generation and deep neural network. Given an image, object hypotheses are generated in an unsupervised manner. Then we extract the image features for each hypothesis with a deep neural network model. By combining the features of all hypotheses, we get the features of the entire image. Finally, we calculate for each label the probability of that the label is correlated with the given image. It can be trained in an end-to-end way using the standard backward propagation algorithm. Experimental results on multiple benchmark datasets show that our method is better than the state-of-the-art ones.