Yixin Chen

MILES: Multiple-Instance Learning via Embedded Instance Selection

Multiple-instance problems arise from the situations where training class labels are attached to sets of samples (named bags), instead of individual samples within each bag (called instances). Most previous multiple-instance learning (MIL) algorithms are developed based on the assumption that a bag is positive if and only if at least one of its instances is positive. Although the assumption works well in a drug activity prediction problem, it is rather restrictive for other applications, especially those in the computer vision area. We propose a learning method, MILES (multiple-instance learning via embedded instance selection), which converts the multiple-instance learning problem to a standard supervised learning problem that does not impose the assumption relating instance labels to bag labels. MILES maps each bag into a feature space defined by the instances in the training bags via an instance similarity measure. This feature mapping often provides a large number of redundant or irrelevant features. Hence, 1-norm SVM is applied to select important features as well as construct classifiers simultaneously. We have performed extensive experiments. In comparison with other methods, MILES demonstrates competitive classification accuracy, high computation efficiency, and robustness to labeling uncertainty

A region-based fuzzy feature matching approach to content-based image retrieval

This paper proposes a fuzzy logic approach, UFM (unified feature matching), for region-based image retrieval. In our retrieval system, an image is represented by a set of segmented regions, each of which is characterized by a fuzzy feature (fuzzy set) reflecting color, texture, and shape properties. As a result, an image is associated with a family of fuzzy features corresponding to regions. Fuzzy features naturally characterize the gradual transition between regions (blurry boundaries) within an image and incorporate the segmentation-related uncertainties into the retrieval algorithm. The resemblance of two images is then defined as the overall similarity between two families of fuzzy features and quantified by a similarity measure, UFM measure, which integrates properties of all the regions in the images. Compared with similarity measures based on individual regions and on all regions with crisp-valued feature representations, the UFM measure greatly reduces the influence of inaccurate segmentation and provides a very intuitive quantification. The UFM has been implemented as a part of our experimental SIMPLIcity image retrieval system. The performance of the system is illustrated using examples from an image database of about 60,000 general-purpose images.

CLUE: cluster-based retrieval of images by unsupervised learning

In a typical content-based image retrieval (CBIR) system, target images (images in the database) are sorted by feature similarities with respect to the query. Similarities among target images are usually ignored. This paper introduces a new technique, cluster-based retrieval of images by unsupervised learning (CLUE), for improving user interaction with image retrieval systems by fully exploiting the similarity information. CLUE retrieves image clusters by applying a graph-theoretic clustering algorithm to a collection of images in the vicinity of the query. Clustering in CLUE is dynamic. In particular, clusters formed depend on which images are retrieved in response to the query. CLUE can be combined with any real-valued symmetric similarity measure (metric or nonmetric). Thus, it may be embedded in many current CBIR systems, including relevance feedback systems. The performance of an experimental image retrieval system using CLUE is evaluated on a database of around 60,000 images from COREL. Empirical results demonstrate improved performance compared with a CBIR system using the same image similarity measure. In addition, results on images returned by Googles Image Search reveal the potential of applying CLUE to real-world image data and integrating CLUE as a part of the interface for keyword-based image retrieval systems.

Content-based image retrieval by clustering

In a typical content-based image retrieval (CBIR) system, query results are a set of images sorted by feature similarities with respect to the query. However, images with high feature similarities to the query may be very different from the query in terms of semantics. This is known as the semantic gap. We introduce a novel image retrieval scheme, CLUster-based rEtrieval of images by unsupervised learning (CLUE), which tackles the semantic gap problem based on a hypothesis: semantically similar images tend to be clustered in some feature space. CLUE attempts to capture semantic concepts by learning the way that images of the same semantics are similar and retrieving image clusters instead of a set of ordered images. Clustering in CLUE is dynamic. In particular, clusters formed depend on which images are retrieved in response to the query. Therefore, the clusters give the algorithm as well as the users semantic relevant clues as to where to navigate. CLUE is a general approach that can be combined with any real-valued symmetric similarity measure (metric or nonmetric). Thus it may be embedded in many current CBIR systems. Experimental results based on a database of about 60, 000 images from COREL demonstrate improved performance.

A sparse support vector machine approach to region-based image categorization

Automatic image categorization using low-level features is a challenging research topic in computer vision. In this paper, we formulate the image categorization problem as a multiple-instance learning (MIL) problem by viewing an image as a bag of instances, each corresponding to a region obtained from image segmentation. We propose a new solution to the resulting MIL problem. Unlike many existing MIL approaches that rely on the diverse density framework, our approach performs an effective feature mapping through a chosen metric distance function. Thus the MIL problem becomes solvable by a regular classification algorithm. Sparse SVM is adopted to dramatically reduce the regions that are needed to classify images. The selected regions by a sparse SVM approximate to the target concepts in the traditional diverse density framework. The proposed approach is a lot more efficient in computation and less sensitive to the class label uncertainty. Experimental results are included to demonstrate the effectiveness and robustness of the proposed method.

Kernel machines and additive fuzzy systems: classification and function approximation

This paper investigates the connection between additive fuzzy systems and kernel machines. We prove that, under quite general conditions, these two seemingly quite distinct models are essentially equivalent. As a result, algorithms based upon support vector (SV) learning are proposed to build fuzzy systems for classification and function approximation. The performance of the proposed algorithm is illustrated using extensive experimental results.

A content-based image retrieval system for fish taxonomy

It is estimated that less than ten percent of the worlds species have been discovered and described. The main reason for the slow pace of new species description is that the science of taxonomy, as traditionally practiced, can be very laborious: taxonomists have to manually gather and analyze data from large numbers of specimens, often from broad geographic areas, and identify the smallest subset of external body characters that uniquely diagnoses the new species as distinct from all its known relatives. The pace of data gathering and analysis in taxonomy can be greatly increased by the development of information technology. The Internet is being used to link taxonomists,taxonomic literature and specimen databases in different parts of the globe, and hence enables the development of tools for remote study of specimens archived as digital images. In this paper, we propose a content-based image retrieval system for taxonomic research. The system has a learning component that can identify representative body shape characters of known species based on digitized landmarks. The system can also provide statistical clues for assisting taxonomists to identify new species or subspecies. The experiments on a taxonomic problem involving species of suckers in the genera Carpiodes demonstrate promising results.

FIRM: fuzzily integrated region matching for content-based image retrieval

We propose FIRM (Fuzzily Integrated Region Matching), an efficient and robust similarity measure for region-based image retrieval. Each image in our retrieval system is represented by a set of regions that are characterized by fuzzy sets. The FIRM measure, representing the overall similarity between two images, is defined as the similarity between two families of fuzzy sets. Compared with similarity measures based on individual regions and on all regions with crisp feature representations, our approach greatly reduces the influence of inaccurate segmentation. Experimental results based on a database of about 200,000 general-purpose images demonstrate improved accuracy, robustness, and high speed.

Machine annotation and retrieval for digital imagery of historical materials

Annotating digital imagery of historical materials for the purpose of computer-based retrieval is a labor-intensive task for many historians and digital collection managers. We have explored the possibilities of automated annotation and retrieval of images from collections of art and cultural images. In this paper, we introduce the application of the ALIP (Automatic Linguistic Indexing of Pictures) system, developed at Penn State, to the problem of machine-assisted annotation of images of historical materials. The ALIP system learns the expertise of a human annotator on the basis of a small collection of annotated representative images. The learned knowledge about the domain-specific concepts is stored as a dictionary of statistical models in a computer-based knowledge base. When an un-annotated image is presented to ALIP, the system computes the statistical likelihood of the image resembling each of the learned statistical models and the best concept is selected to annotate the image. Experimental results, obtained using the Emperor image collection of the Chinese Memory Net project, are reported and discussed. The system has been trained using subsets of images and metadata from the Emperor collection. Finally, we introduce an integration of wavelet-based annotation and wavelet-based progressive displaying of very high resolution copyright-protected images.

Looking beyond region boundaries: a robust image similarity measure using fuzzified region features

The performance of most region-based image retrieval systems depend critically on the accuracy of object segmentation. We propose a region matching approach, unified feature matching (UFM), which greatly increases the robustness of the retrieval system against segmentation related uncertainties. In our retrieval system, an image is represented by a set of segmented regions each of which is characterized by a fuzzy feature reflecting color, texture, and shape properties. The resemblance between two images is then defined as the overall similarity between two families of fuzzy features, and quantified by the UFM measure. The system has been tested on a database of about 60,000 general-purpose images. Experimental results demonstrate improved accuracy and robustness.