Conference Paper

Semantic manifold learning for image retrieval.

DOI: 10.1145/1101149.1101193 Conference: Proceedings of the 13th ACM International Conference on Multimedia, Singapore, November 6-11, 2005
Source: DBLP

ABSTRACT Learning the user's semantics for CBIR involves two different sources of information: the similarity relations entailed by the content-based features, and the relevance relations specified in the feedback. Given that, we propose an augmented relation embedding (ARE) to map the image space into a semantic manifold that faithfully grasps the user's preferences. Besides ARE, we also look into the issues of selecting a good feature set for improving the retrieval performance. With these two aspects of efforts we have established a system that yields far better results than those previously reported. Overall, our approach can be characterized by three key properties: 1) The framework uses one relational graph to describe the similarity relations, and the other two to encode the relevant/irrelevant relations indicated in the feedback. 2) With the relational graphs so defined, learning a semantic manifold can be transformed into solving a constrained optimization problem, and is reduced to the ARE algorithm accounting for both the representation and the classification points of views. 3) An image representation based on augmented features is introduced to couple with the ARE learning. The use of these features is significant in capturing the semantics concerning different scales of image regions. We conclude with experimental results and comparisons to demonstrate the effectiveness of our method.

  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we propose a novel semi-supervised feature selection framework by mining correlations among multiple tasks and apply it to different multimedia applications. Instead of independently computing the importance of features for each task, our algorithm leverages shared knowledge from multiple related tasks, thus, improving the performance of feature selection. Note that we build our algorithm on assumption that different tasks share common structures. The proposed algorithm selects features in a batch mode, by which the correlations between different features are taken into consideration. Besides, considering the fact that labeling a large amount of training data in real world is both time-consuming and tedious, we adopt manifold learning which exploits both labeled and unlabeled training data for feature space analysis. Since the objective function is non-smooth and difficult to solve, we propose an iterative algorithm with fast convergence. Extensive experiments on different applications demonstrate that our algorithm outperforms other state-of-the-art feature selection algorithms.
  • [Show abstract] [Hide abstract]
    ABSTRACT: Discriminant neighborhood embedding (DNE) is a typical graph-based dimensionality reduction method, and has been successfully applied to face recognition. By constructing an adjacency graph, aiming to keep the local structure for original data in the subspace, it is able to find the optimal discriminant direction effectively. Not for every sample does DNE set up a link between it and its heterogeneous samples when constructing the adjacency graph, which would result in a small between-class scatter. Motivated by this fact, we develop an extension of DNE, called double adjacency graphs-based discriminant neighborhood embedding (DAG-DNE) by introducing two adjacency graphs, or homogeneous and heterogeneous neighbor adjacency graphs. In DAG-DNE, neighbors belonging to the same class are compact while neighbors belonging to different classes become separable in the subspace. Thus, DAG-DNE could keep the local structure of a given data and find a good projection matrix for them. To investigate the performance of DAG-DNE, we compare it with the state-of-the-art dimensionality reduction techniques such as DNE and MFA on several publicly available datasets. Experimental results show the feasibility and effectiveness of the proposed method.
    Pattern Recognition 05/2015; 48(5). DOI:10.1016/j.patcog.2014.08.025 · 2.58 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Manifold learning is a well-known dimensionality reduction scheme which can detect intrinsic low-dimensional structures in non-linear high-dimensional data. It has been recently widely employed in data analysis, pattern recognition, and machine learning applications. Isomap is one of the most promising manifold learning algorithms, which extends metric multi-dimensional scaling by using approximate geodesic distance. However, when Isomap is conducted on real-world applications, it may have some difficulties in dealing with noisy data. Although many applications represent a special sample by multiple feature vectors in different spaces, Isomap employs samples in unique observation space. In this paper, two extended versions of Isomap to multiple feature spaces problem, namely fusion of dissimilarities and fusion of geodesic distances, are presented. We have employed the advantages of several spaces and depicted the Euclidean distance on learned manifold that is more compatible to the semantic distance. To show the effectiveness and validity of the proposed method, some experiments have been carried out on the application of shape analysis on MPEG7 CE Part B and Fish data sets.
    Machine Vision and Applications 05/2014; 25(4):1053-1065. DOI:10.1007/s00138-014-0604-7 · 1.44 Impact Factor

Full-text (2 Sources)

Available from
Jun 3, 2014