Topic Models for Image Annotation and Text Illustration
نویسندگان
چکیده
Image annotation, the task of automatically generating description words for a picture, is a key component in various image search and retrieval applications. Creating image databases for model development is, however, costly and time consuming, since the keywords must be hand-coded and the process repeated for new collections. In this work we exploit the vast resource of images and documents available on the web for developing image annotation models without any human involvement. We describe a probabilistic model based on the assumption that images and their co-occurring textual data are generated by mixtures of latent topics. We show that this model outperforms previously proposed approaches when applied to image annotation and the related task of text illustration despite the noisy nature of our dataset.
منابع مشابه
Semantic-Based Image Retrial in the VQ Compressed Domain using Image Annotation Statistical Models
متن کامل
Image Annotation and Two Paths to Text Illustration
This paper describes our participation to the ImageCLEF 2016 scalable concept image annotation main task and Text Illustration teaser. Regarding image annotation, we focused on better localizing the detected features. For this, we identified the saliency of the image to collect a list of potential interesting places into the image. We also added a specific human attribute detector that boosted ...
متن کاملBayesian Models for Massive Multimedia Databases: a New Frontier
Modelling the increasing number of digital databases (the web, photo-libraries, music collections, news archives, medical databases) is one of the greatest challenges of statisticians in the new century. Despite the large amounts of data, the models are so large that they motivate the use of Bayesian models. In particular, the Bayesian perspective allows us to perform automatic regularisation t...
متن کاملFuzzy Neighbor Voting for Automatic Image Annotation
With quick development of digital images and the availability of imaging tools, massive amounts of images are created. Therefore, efficient management and suitable retrieval, especially by computers, is one of themost challenging fields in image processing. Automatic image annotation (AIA) or refers to attaching words, keywords or comments to an image or to a selected part of it. In this paper,...
متن کاملTags Re-ranking Using Multi-level Features in Automatic Image Annotation
Automatic image annotation is a process in which computer systems automatically assign the textual tags related with visual content to a query image. In most cases, inappropriate tags generated by the users as well as the images without any tags among the challenges available in this field have a negative effect on the query's result. In this paper, a new method is presented for automatic image...
متن کامل