Generic image classification using visual knowledge on the web

Abstract

In this paper, we describe a generic image classification system with an automatic knowledge acquisition mechanism from the World Wide Web. Due to the recent spread of digital imaging devices, the demand for image recognition of various kinds of real world scenes becomes greater. For realizing it, visual knowledge on various kinds of scenes is required. Then, we propose gathering visual knowledge on real world scenes for generic image classification from the World Wide Web. Our system gathers a large number of images from the Web automatically and makes use of them as training images for generic image classification. It consists of three modules, which are an image-gathering module, an image-learning module and an image classification module. The image-gathering module gathers images related to given class keywords from the Web automatically. The learning module extracts image features from gathered images and associates them with each class. The image classification module classifies an unknown image into one of the classes corresponding to the class keywords by using the association between image features and classes. In the experiments, the system demonstrated potential for generic image classification/recognition using images gathered from the World Wide Web automatically as training images.

A fast image-gathering system from the World-Wide Web using a PC cluster

“…Furthermore, recently we are studying a Web image mining project [7], which focuses on a generic image recognition/classification by using gathered images from WWW as training images. For that project, we are extending this system in order to gather much more Web images.…”

“…Our initial motivation for implementing this system is to gather a large number of images for 'Web image mining' research [7]. The ability of gathering a large number of images automatically is an important property for Web image mining.…”

“…The property without a user's intervention during the processing is important for us, since our objective of this system is gathering a large number of images related given keywords. Our initial motivation for implementing this system is to gather a large number of images for 'Web image mining' research [7].…”

Event photo mining from Twitter using keyword bursts and image clustering

“…In this paper, we perform event photo clustering for image selection as well as visualization. Instead of k-means which requires the given number of clusters in advance, we use hierarchical clustering to select relevant clusters, which is the same as our past work on Web image selection [23]. Because hierarchical clustering needs no fixed number of clusters, it was commonly used in work on Web image clustering [24].…”

Object Recognition by Integrated Information Using Web Images

“…There are some works of the image recognition using Web images [8], [9], [10]. In [8], large amount of Web images are collected and the reranking of the images are performed in order to remove the noise. However, considering real-time learning, the recognition using the small amount of images is needed.…”