LightBlog
Contact at mumbai.academics@gmail.com or 8097636691/9323040215
Responsive Ads Here

Monday, 26 February 2018

An Attribute-Assisted Reranking Model for Web Image Search


An Attribute-Assisted Reranking 

Model for Web Image Search


Image search reranking is an effective approach to refine the text-based image search result. Most existing reranking approaches are based on low-level visual features. In this paper, we propose to exploit semantic attributes for image search reranking. Based on the classifiers for all the predefined attributes, each image is represented by an attribute feature consisting of the responses from these classifiers. A hypergraph is then used to model the relationship between images by integrating low-level visual features and attribute features. Hypergraph ranking is then performed to order the images. Its basic principle is that visually similar images should have similar ranking scores. In this paper, we propose a visual-attribute joint hypergraph learning approach to simultaneously explore two information sources. A hypergraph is constructed to model the relationship of all images. We conduct experiments on more than 1,000 queries in MSRA-MMV2.0 data set. The experimental results demonstrate the effectiveness of our approach.An Attribute-Assisted Reranking Model for Web Image Search
PROPOSED SYSTEM:
  • We propose a new attribute-assisted reranking method based on hypergraph learning. We first train several classifiers for all the pre-defined attributes and each image is represented by attribute feature consisting of the responses from these classifiers.
  • We improve the hypergraph learning method approach by adding a regularizer on the hyperedge weights which performs an implicit selection on the semantic attributes.
  • This paper serves as a first attempt to include the attributes in reranking framework. We observe that semantic attributes are expected to narrow down the semantic gap between low-level visual features and high level semantic meanings.
ADVANTAGES OF PROPOSED SYSTEM:
  • We propose a novel attribute-assisted retrieval model for reranking images. Based on the classifiers for all the predefined attributes.
  • We perform hypergraph ranking to re-order the images, which is also constructed to model the relationship of all images.
  • Our proposed iterative regularization framework could further explore the semantic similarity between images by aggregating their local
  • Compared with the previous method, a hypergraph is reconstructed to model the relationship of all the images, in which each vertex denotes an image and a hyperedge represents an attribute and a hyperedge connects to multiple vertices.

No comments:

Post a Comment