Indexing Text and Visual Features for WWW Images

Abstract

In this paper, we present a novel indexing technique called Multi-scale Similarity Indexing (MSI) to index image's multi-features into a single one-dimensional structure. Both for text and visual feature spaces, the similarity between a point and a local partition's center in individual space is used as the indexing key, where similarity values in different features are distinguished by different scale. Then a single indexing tree can be built on these keys. Based on the property that relevant images haves similar similarity values from the center of the same local partition in any feature space, certain number of irrelevant images can be fast pruned based on the triangle inequity on indexing keys. To remove the "dimensionality curse" existing in high dimensional structure, we propose a new technique called Local Bit Stream (LBS). LBS transforms image's text and visual feature representations into simple, uniform and effective bit stream (BS) representations based on local partition's center. Such BS representations are small in size and fast for comparison since only bit operation are involved. By comparing common bits existing in two BSs, most of irrelevant images can be immediately filtered. Our extensive experiment showed that single one-dimensional index on multi-features improves multi-indices on multi-features greatly. Our LBS method outperforms sequential scan on high dimensional space by an order of magnitude.

Speed up kernel discriminant analysis

“…Once the high-dimensional data is mapped into lowerdimensional space, conventional indexing schemes can then be applied [20,28,29].…”

Hybrid index-based image search from the web

“…Indexing methods to search for images and videos with both text and content information were proposed in Smith and Chang (1996) and Smith et al (2001) but they do not support simultaneous querying using hybrid query feature vectors. In Shen et al (2005Shen et al ( , 2006, a multi-scale similarity indexing method was proposed for indexing both text and content features of images. It tries to reduce all features into a one dimensional key space and uses standard B+ tree to index the keys.…”

A Latent Image Semantic Indexing Scheme for Image Retrieval on the Web

“…As wavelet transformation is a powerful tool in effectively generating compact representation of visual features of images [10], we adopt the Daubechies' wavelets to extract the wavelet coefficients using the LUV color space for its good perception correlation properties [15,7]. The feature elements chosen from the wavelet coefficients are independent of image resolution and scaling.…”

“…Recent research efforts in CBIR focuses on bridging the so-called "semantic gap" which combines Relevance Feedback techniques or text and keyword predicates to get powerful retrieval methods for image collections [3,9,10]. Users intend to specify the semantic class of the scene or the objects it should contain when they search for images.…”