In many vision applications, the practice of supervised learning faces several difficulties, one of which is that insufficient labeled training data result in poor generalization. In image retrieval, we have very few labeled images from query and relevance feedback so that it is hard to automatically weight image features and select similarity metrics for image classification. This paper investigates the possibility of including an unlabeled data set to make up the insufficiency of labeled data. Different from most current research in image retrieval, the proposed approach tries to cast image retrieval as a transductive learning problem, in which the generalization of an image classifier is only defined on a set of images such as the given image database. Formulating this transductive problem in a probabilistic framework, the proposed algorithm, Discriminant-EM (D-EM), not only estimates the parameters of a generative model, but also finds a linear transformation to relax the assumption of pro...

AbstractÐThe main motivation of this paper is to propose a new classification and regression method for challenging highdimensional data. The proposed new technique casts classification problems (class labels as output) and regression problems (numeric values as output) into a unified regression problem. This unified view enables classification problems to use numeric information in the output space that is available for regression problems but are traditionally not readily available for classification problemsÐdistance metric among clustered class labels for coarse and fine classifications. A doubly clustered subspace-based hierarchical discriminating regression (HDR) method is proposed in this work. The major characteristics include: 1) Clustering is performed in both output space and input space at each internal node, termed ªdoubly clustered.º Clustering in the output space provides virtual labels for computing clusters in the input space. 2) Discriminants in the input space are automatically derived from the clusters in the input space. These discriminants span the discriminating subspace at each internal node of the tree. 3) A hierarchical probability distribution model is applied to the resulting discriminating subspace at each internal node. This realizes a coarse-to-fine approximation of probability distribution of the input samples, in the hierarchical discriminating subspaces. No global distribution models are assumed. 4) To relax the per class sample requirement of traditional discriminant analysis techniques, a sample-size dependent negative-log-likelihood (NLL) is introduced. This new technique is designed for automatically dealing with small-sample applications, large-sample applications, and unbalanced-sample applications. 5) The execution of HDR method is fast, due to the empirical logarithmic time complexity of the HDR algorithm. Although the method is applicable to any data, we report the experimental results for three types of data: synthetic data for examining the near-optimal performance, large raw face-image data bases, and traditional databases with manually selected features along with a comparison with some major existing methods, such as CART,

Content-based image retrieval (CBIR) has become one of the most active research areas in the past few years. Most of the attention from the research has been focused on indexing techniques based on global feature distributions. However, these global distributions have limited discriminating power because they are unable to capture local image information. The use of interest points in content-based image retrieval allow image index to represent local properties of the image. Classic corner detectors can be used for this purpose. However, they have drawbacks when applied to various natural images for image retrieval, because visual features need not be corners and corners may gather in small regions. In this paper, we present a salient point detector. The detector is based on wavelet transform to detect global variations as well as local ones. The wavelet-based salient points are evaluated for image retrieval with a retrieval system using color and texture features. The results show that salient points with Gabor feature perform better than the other point detectors from the literature and the randomly chosen points. Significant improvements are achieved in terms of retrieval accuracy, computational complexity when compared to the global feature approaches. 2001 SPIE and IS&T. [DOI: 10.1117/1.1406945] 1

. It is desirable for humans to control humanoids through high-level commands, but it is too tedious for humans to issue commands for every detailed action for every fraction of a second. However, it is extremely challenging for humans to program a humanoid robot to such a sufficient degree that it acts properly in typical unknown human environments. This is especially true for humanoids due to the very large number of redundant degrees of freedom and a large number of sensors that are required for humanoids to work safely and effectively in the human environment. How can we address this fundamental problem? Motivated by human mental development from infancy to adulthood, we enable robots develop its mind automatically, through online, real time interactions with its environment. Humans mentally "raise" the robot through "robot sitting" and "robot schools" instead of task-specific robot programming. The SAIL developmental robot that has been built at MSU is an early prototype ...

Abstract not found

We present a user-centric system for visualization and layout for content-based image retrieval. Image features (visual and/or semantic) are used to display retrievals as thumbnails in a 2-D spatial layout or "configuration" which conveys all pair-wise mutual similarities. A graphical optimization technique is used to provide maximally uncluttered and informative layouts. Moreover, a novel subspace feature weighting technique can be used to modify 2-D layouts in a variety of context-dependent ways. An efficient computational technique for subspace weighting and re-estimation leads to a simple user-modeling framework whereby the system can learn to display query results based on layout examples (or relevance feedback) as provided by the user. The resulting retrieval, browsing and visualization engine can adapt to the users' (time-varying) notions of content, context and preferences in presentation style and interactive navigation. Monte Carlo simulations with machine-generated layouts as well as pilot user studies have demonstrated the ability of this framework to model or "mimic" users, by automatically generating layouts according to their preferences.

The Non-negative Matrix Factorization technique (NMF) has been recently proposed for dimensionality reduction. NMF is capable to produce a region- or partbased representation of objects and images. The positive space defined with NMF lacks of a suitable metric and this paper experimentally compares NMF to Principal Component Analysis (PCA) in the context of classification trying to determine the best distance metric for the NMF. This paper introduces the use of the Earth Mover's Distance (EMD) as a relevant metric that takes into account the positive definition of the NMF bases leading to obtain the best recognition results when the dimensionality of the problem is correctly chosen. PCA and NMF have also been tested under the presence of occlusions and due to its part-based representation, NMF is able to deal with occlusions improving the PCA results.

This paper presents Incremental Hierarchical Discriminant Regression (IHDR) which incrementally builds a decision tree or regression tree for very high dimensional regression or decision spaces by an online, real-time learning system. Biologically motivated, it is an approximate computational model for automatic development of associative cortex, with both bottom-up sensory inputs and top-down motor projections. At each internal node of the IHDR tree, information in the output space is used to automatically derive the local subspace spanned by the most discriminating features. Embedded in the tree is a hierarchical probability distribution model used to prune very unlikely cases during the search. The number of parameters in the coarse-to-fine approximation is dynamic and data-driven, enabling the IHDR tree to automatically fit data with unknown distribution shapes (thus, it is difficult to select the number of parameters up front). The IHDR tree dynamically assigns long-term memory to avoid the loss-of-memory problem typical with a global-fitting learning algorithm for neural networks. A major challenge for an incrementally built tree is that the number of samples varies arbitrarily during the construction process. An incrementally updated probability model, called sample size dependent negative-log-likelihood (SDNLL) metric is used to deal with large-sample size cases, small-sample size cases, and unbalanced-sample size cases, measured among different internal nodes of the IHDR tree. We report experimental results for four types of data: synthetic data to visualize the behavior of the algorithms, large face image data, continuous video stream from robot navigation, and publicly available data sets that use human defined features.

Retrieval of images from large image archives based solely on their visual similarity to a query image provides an exciting alternative to conventional text-based search. For content-based retrieval images are represented in terms of visual features. The question of how to combine these for similarity computation is typically addressed by eliciting relevance feedback from the user on the retrieved images. We argue in this thesis that the prevailing approach to relevance feedback suffers from three significant shortcomings: firstly, it leaves unsolved the question of how to combine features for the first retrieval; secondly, the advantage of automated content-extraction over manual annotation is greatest for large collections but if the query image is not constrained to come from the indexed collection, content-based retrieval entails imagewise comparisons leading to prohibitive response times; thirdly, users may only have vaguely defined information needs or may change their needs in the course of the interaction. The large majority of relevance feedback techniques are ill-suited for such undirected exploration. We propose a new framework of user interaction that addresses these limitations. It is centred on what we call the NN k idea. The NN k of an image are all those images that are most similar to it under some combination of features. They can be viewed as representatives of the possible

Abstract—We introduce a methodology for adaptive sequential sensing and processing in a classification setting. Our objective for sensor optimization is the back-end performance metric—in this case, misclassification rate. Our methodology, which we dub Integrated Sensing and Processing Decision Trees (ISPDT), optimizes adaptive sequential sensing for scenarios in which sensor and/or throughput constraints dictate that only a small subset of all measurable attributes can be measured at any one time. Our decision trees optimize misclassification rate by invoking a local dimensionality reduction-based partitioning metric in the early stages, focusing on classification only in the leaves of the tree. We present the ISPDT methodology and illustrative theoretical, simulation, and experimental results. Index Terms—Classification, clustering, adaptive sensing, sequential sensing, local dimensionality reduction. æ 1