Multichannel Texture Analysis Using Localized Spatial Filters
IEEE Transactions on Pattern Analysis and Machine Intelligence
Query refinement for multimedia similarity retrieval in MARS
MULTIMEDIA '99 Proceedings of the seventh ACM international conference on Multimedia (Part 1)
Content-Based Image Retrieval at the End of the Early Years
IEEE Transactions on Pattern Analysis and Machine Intelligence
Matrix algorithms
Introduction to Modern Information Retrieval
Introduction to Modern Information Retrieval
Object Recognition as Machine Translation: Learning a Lexicon for a Fixed Image Vocabulary
ECCV '02 Proceedings of the 7th European Conference on Computer Vision-Part IV
MindReader: Querying Databases Through Multiple Examples
VLDB '98 Proceedings of the 24rd International Conference on Very Large Data Bases
IEEE Transactions on Pattern Analysis and Machine Intelligence
Semantic manifold learning for image retrieval
Proceedings of the 13th annual ACM international conference on Multimedia
Graph Embedding and Extensions: A General Framework for Dimensionality Reduction
IEEE Transactions on Pattern Analysis and Machine Intelligence
A tutorial on spectral clustering
Statistics and Computing
Spectral regression: a unified subspace learning framework for content-based image retrieval
Proceedings of the 15th international conference on Multimedia
Learning a Maximum Margin Subspace for Image Retrieval
IEEE Transactions on Knowledge and Data Engineering
| Hi-index | 0.00 |
The "semantic gap" problem is one of the main difficulties in image retrieval task. Semi-supervised learning is an effective methodology proposed to narrow down the gap, which is also often integrated with relevance feedback techniques. However, in semi-supervised learning, the amount of unlabeled data is usually much greater than that of labeled data. Therefore, the performance of a semi-supervised learning algorithm relies heavily on how effective it uses the relationship between the labeled and unlabeled data. A novel algorithm is proposed in this paper to enhance the relational graph built on the entire data set, expected to increase the intra-class weights of data while decreasing the inter-class weights and linking the potential intra-class data. The enhanced relational matrix can be directly used in any semi-supervised learning algorithm. The experimental results in feedback-based image retrieval tasks show that the proposed algorithm performs much better compared with other algorithms in the same semi-supervised learning framework.