EFFICIENTLY IDENTIFYING IMAGES, VIDEOS, SONGS OR DOCUMENTS MOST RELEVANT TO THE USER BASED ON ATTRIBUTE FEEDBACK

摘要

A method, system and computer program product for efficiently identifying images, videos, audio files or documents relevant to a user. Using either manual annotations or learned functions, the method predicts the relative strength of an attribute in an image, video, audio file or document from a pool of images, videos, audio files or documents. At query time, the system presents an initial set of reference images, videos, audio files or documents, and the user selects among them to provide relative attribute feedback. Using the resulting constraints in the multi-dimensional attribute space, the relevance function for the pool of images, videos, audio files or documents is updated and the relevance of the pool of images, videos, audio files or documents is re-computed. This procedure iterates using the accumulated constraints until the top-ranked images, videos, audio files or documents are acceptably close to the user's envisioned image, video, audio file or document.

主权项

1. A method for efficiently identifying images, videos, audio files or documents relevant to a user, the method comprising:
determining a set of ranking functions, each of which predicts or assigns a relative strength of an attribute in an image, video, audio file or document from a pool of database images, videos, audio files or documents; presenting a set of reference images, videos, audio files or documents from said pool of database images, videos, audio files or documents; receiving a selection of one or more images, videos, audio files or documents from said set of reference images, videos, audio files or documents along with relative attribute feedback pertaining to one or more attributes of said selected one or more images, videos, audio files or documents; and revising, by a processor, a system's model of what images, videos, audio files or documents are relevant to said user using said relative attribute feedback.