Transfer incremental learning for pattern classification.

SKIF a data imputation framework for concept drifting data streams.

Ensemble pruning via individual contribution ordering.
An ensemble is a set of learned models that make decisions collectively. Although an ensemble is usually more accurate than a single learner, existing ensemble methods often tend to construct unnecessarily large ensembles, which increases the memory consumption and computational cost. Ensemble pruning tackles this problem by selecting a subset of ensemble members to form subensembles that are subject to less resource consumption and response time with accuracy that is similar to or better than the original ensemble. In this paper, we analyze the accuracy/diversity trade-off and prove that classifiers that are more accurate and make more predictions in the minority group are more important for subensemble construction. Based on the gained insights, a heuristic metric that considers both accuracy and diversity is proposed to explicitly evaluate each individual classifier’s contribution to the whole ensemble. By incorporating ensemble members in decreasing order of their contributions, subensembles are formed such that users can select the top p percent of ensemble members, depending on their resource availability and tolerable waiting time, for predictions. Experimental results on 26 UCI data sets show that subensembles formed by the proposed EPIC (Ensemble Pruning via Individual Contribution ordering) algorithm outperform the original ensemble and a state-ofthe-art ensemble pruning method, Orientation Ordering (OO) [16].
Rule Synthesizing from Multiple Related Databases.

Feature Selection for Datasets with Imbalanced Class Distributions.

Feature Selection for Datasets with Imbalanced Class Distributions.

Normalized dimensionality reduction using nonnegative matrix factorization.

Multiple criteria programming models for VIP E%2DMail behavior analysis.

The Impact of Gene Selection on Imbalanced Microarray Expression Data.

Bias%2DVariance Analysis for Ensembling Regularized Multiple Criteria Linear Programming Models.

Mining Data Streams with Labeled and Unlabeled Training Examples.

Vague One%2DClass Learning for Data Streams.

Approximate Repeating Pattern Mining with Gap Requirements.

Error Detection and Uncertainty Modeling for Imprecise Data.

Multiple Information Sources Cooperative Learning.
Many applications are facing the problem of learning from an objective dataset, whereas information from other auxiliary sources may be beneficial but cannot be integrated into the objective dataset for learning. In this paper, we propose an omni-view learning approach to enable learning from multiple data collections. The theme is to organize heterogeneous data sources into a unified table with global data view. To achieve the omni-view learning goal, we consider that the objective dataset and the auxiliary datasets share some instance-level dependency structures. We then propose a relational k-means to cluster instances in each auxiliary dataset, such that clusters can help build new features to capture correlations between the objective and auxiliary datasets. Experimental results demonstrate that omni-view learning can help build models which outperform the ones learned from the objective dataset only. Comparisons with the co-training algorithm further assert that omni-view learning provides an alternative, yet effective, way for semi-supervised learning. 1
An Aggregate Ensemble for Mining Concept Drifting Data Streams with Noise.

Feature Selection with biased Sample Distributions.

An Aggregate Ensemble for Mining Concept Drifting Data Streams with Noise.

Quantitative Association Rules.
 
vEye behavioral footprinting for self%2Dpropagating worm detection and profiling.

Contrast Pattern Mining with Gap Constraints for Peptide Folding Prediction.

Cleansing Noisy Data Streams.



VoB predictors Voting on bagging classifications.

VCI predictors Voting on classifications from imputed learning sets.

An empirical study of supervised learning for biological sequence profiling and microarray expression data analysis.

Categorizing and mining concept drifting data streams.

iVESTA an interactive visualization and evaluation system for drive test data.

Imputation%2Dboosted collaborative filtering using machine learning classifiers.

Imputation%2Dboosted collaborative filtering using machine learning classifiers.

Select Objective Functions for Multiple Criteria Programming Classification.

Data Mining.

Conceptual equivalence for contrast mining in classification learning.
Learning often occurs through comparing. In classification learning, in order to compare data groups, most existing methods compare either raw instances or learned classification rules against each other. This paper takes a different approach, namely conceptual equivalence, that is, groups are equivalent if their underlying concepts are equivalent while their instance spaces do not necessarily overlap and their rule sets do not necessarily present the same appearance. A new methodology of comparing is proposed that learns a representation of each group’s underlying concept and respectively crossexams one group’s instances by the other group’s concept representation. The innovation is five-fold. First, it is able to quantify the degree of conceptual equivalence between two groups. Second, it is able to retrace the source of discrepancy at two levels: an abstract level of underlying concepts and a specific level of instances. Third, it applies to numeric data as well as categorical data. Fourth, it circumvents direct comparisons between (possibly alarge number of) rules that demand substantial effort. Fifth, it reduces dependency on the accuracy of employed classification algorithms. Empirical evidence suggests that this new methodology is effective and yet simple to use in scenarios such as noise cleansing and concept-change learning.
Mining With Noise Knowledge Error%2DAware Data Mining.

A lazy bagging approach to classification.
In this paper, we propose lazy bagging (LB), which builds bootstrap replicate bags based on the characteristics of test instances. Upon receiving a test instance xk, LB trims bootstrap bags by taking into consideration xk’s nearest neighbors in the training data. Our hypothesis is that an unlabeled instance’s nearest neighbors provide valuable information to enhance local learning and generate a classifier with refined decision boundaries emphasizing the test instance’s surrounding region. In particular, by taking full advantage of xk’s nearest neighbors, classifiers are able to reduce classification bias and variance when classifying xk. As a result, LB, which is built on these classifiers, can significantly reduce classification error, compared with the traditional bagging (TB) approach. To investigate LB’s performance, we first use carefully designed synthetic datasets to gain insight into why LB works and under which conditions it can outperform TB. We then test LB against four rival algorithms on a large suite of 35 real-world benchmark datasets using a variety of statistical tests. Empirical results confirm that LB can statistically significantly outperform alternative methods in terms of reducing classification error.
Editorial.
calls &amp; calendarEDITORIAL
A lazy bagging approach to classification.
In this paper, we propose lazy bagging (LB), which builds bootstrap replicate bags based on the characteristics of test instances. Upon receiving a test instance xk, LB trims bootstrap bags by taking into consideration xk’s nearest neighbors in the training data. Our hypothesis is that an unlabeled instance’s nearest neighbors provide valuable information to enhance local learning and generate a classifier with refined decision boundaries emphasizing the test instance’s surrounding region. In particular, by taking full advantage of xk’s nearest neighbors, classifiers are able to reduce classification bias and variance when classifying xk. As a result, LB, which is built on these classifiers, can significantly reduce classification error, compared with the traditional bagging (TB) approach. To investigate LB’s performance, we first use carefully designed synthetic datasets to gain insight into why LB works and under which conditions it can outperform TB. We then test LB against four rival algorithms on a large suite of 35 real-world benchmark datasets using a variety of statistical tests. Empirical results confirm that LB can statistically significantly outperform alternative methods in terms of reducing classification error.
MMIS07%2C 08 mining multiple information sources workshop report.

MMIS07%2C 08 mining multiple information sources workshop report.

SAIL%2DAPPROX An Efficient On%2DLine Algorithm for Approximate Pattern Matching with Wildcards and Length Constraints.

Discovering Relational Patterns across Multiple Databases.

Active Learning from Data Streams.
In this paper, we address a new research problem on active learning from data streams where data volumes grow continuously and labeling all data is considered expensive and impractical. The objective is to label a small portion of stream data from which a model is derived to predict newly arrived instances as accurate as possible. In order to tackle the challenges raised by data streams ’ dynamic nature, we propose a classifier ensembling based active learning framework which selectively labels instances from data streams to build an accurate classifier. A Minimal Variance principle is introduced to guide instance labeling from data streams. In addition, a weight updating rule is derived to ensure that our instance labeling process can adaptively adjust to dynamic drifting concepts in the data. Experimental results on synthetic and real-world data demonstrate the performances of the proposed efforts in comparison with other simple approaches. * 1.
Lazy Bagging for Classifying Imbalanced Data.

An Empirical Study of the Noise Impact on Cost%2DSensitive Learning.

Mining Complex Patterns across Sequences with Gap Requirements.

Mining Frequent Patterns with Wildcards from Biological Sequences.



Pushing Frequency Constraint to Utility Mining Model.

Using WordNet to Disambiguate Word Senses for Text Classification.
Abstract. In this paper, we propose an automatic text classification method based on word sense disambiguation. We use “hood ” algorithm to remove the word ambiguity so that each word is replaced by its sense in the context. The nearest ancestors of the senses of all the non-stopwords in a give document are selected as the classes for the given document. We apply our algorithm to Brown Corpus. The effectiveness is evaluated by comparing the classification results with the classification results using manual disambiguation offered by Princeton University. Keywords: disambiguation, word sense, text classification, WordNet.
Hybrid Collaborative Filtering Algorithms Using a Mixture of Experts.

Mining maximal frequent itemsets from data streams.

Editorial Special issue on mining low%2Dquality data.

Mining Sequential Patterns across Time Sequences.

Accelerated Kernel Feature Analysis.

Error awareness data mining.

Corrective Classification Classifier Ensembling with Corrective and Diverse Base Learners.

Scalable Representative Instance Selection and Ranking.

Bridging Local and Global Data Cleansing Identifying Class Noise in Large
Classification
Mining in Anticipation for Concept Change Proactive%2DReactive Prediction in Data Streams.
Abstract. Prediction in streaming data is an important activity in the modern society. Two major challenges posed by data streams are (1) the data may grow without limit so that it is difficult to retain a long history of raw data; and (2) the underlying concept of the data may change over time. The novelties of this paper are in four folds. First, it uses a measure of conceptual equivalence to organize the data history into a history of concepts. This contrasts to the common practice that only keeps recent raw data. The concept history is compact while still retains essential information for learning. Second, it learns concept-transition patterns from the concept history and anticipates what the concept will be in the case of a concept change. It then proactively prepares a prediction model for the future change. This contrasts to the conventional methodology that passively waits until the change happens. Third, it incorporates proactive and reactive predictions. If the anticipation turns out to be correct, a proper prediction model can be launched instantly
Class Noise Handling for Effective Cost%2DSensitive Learning by Cost%2DGuided Iterative Classification Filtering.

Efficient string matching with wildcards and length constraints.

Effective classification of noisy data streams with attribute%2Doriented dynamic classifier selection.
Abstract. Recently, mining from data streams has become an important and challenging task for many real-world applications such as credit card fraud protection and sensor networking. One popular solution is to separate stream data into chunks, learn a base classifier from each chunk, and then integrate all base classifiers for effective classification. In this paper, we propose a new dynamic classifier selection (DCS) mechanism to integrate base classifiers for effective mining from data streams. The proposed algorithm dynamically selects a single ”best ” classifier to classify each test instance at run time. Our scheme uses statistical information from attribute values, and uses each attribute to partition the evaluation set into disjoint subsets, followed by a procedure that evaluates the classification accuracy of each base classifier on these subsets. Given a test instance, its attribute values determine the subsets that the similar instances in the evaluation set have constructed, and the classifier with the highest classification accuracy on those subsets is selected to classify the test instance. Experimental results and comparative studies demonstrate the efficiency and efficacy of our method. Such a DCS scheme appears to be promising in mining data streams with dramatic concept drifting or with a significant amount of noise, where the base classifiers are likely conflictive or have low confidence.
Sequential Pattern Mining in Multiple Streams.
In this paper, we deal with mining sequential patterns in multiple data streams. Building on a state-of-the-art sequential pattern mining algorithm PrefixSpan for mining transaction databases, we propose MILE 1, an efficient algorithm to facilitate the mining process. MILE recursively utilizes the knowledge of existing patterns to avoid redundant data scanning, and can therefore effectively speed up the new patterns ’ discovery process. Another unique feature of MILE is that it can incorporate some prior knowledge of the data distribution in data streams into the mining process to further improve the performance. Extensive empirical results show that MILE is significantly faster than PrefixSpan. As MILE consumes more memory than PrefixSpan, we also present a solution to balance the memory usage and time efficiency in memory constrained environments. 1.
ACE An Aggressive Classifier Ensemble with Error Detection

Scalable Inductive Learning on Partitioned Data.

Combining proactive and reactive predictions for data streams.
Mining data streams is important in both science and commerce. Two major challenges are (1) the data may grow without limit so that it is difficult to retain a long history; and (2) the underlying concept of the data may change over time. Different from common practice that keeps recent raw data, this paper uses a measure of conceptual equivalence to organize the data history into a history of concepts. Along the journey of concept change, it identifies new concepts as well as re-appearing ones, and learns transition patterns among concepts to help prediction. Different from conventional methodology that passively waits until the concept changes, this paper incorporates proactive and reactive predictions. In a proactive mode, it anticipates what the new concept will be if a future concept change takes place, and prepares prediction strategies in advance. If the anticipation turns out to be correct, a proper prediction model can be launched instantly upon the concept change. If not, it promptly resorts to a reactive mode: adapting a prediction model to the new data. A system RePro is proposed to implement these new ideas. Experiments compare the system with representative existing prediction methods on various benchmark data sets that represent diversified scenarios of concept change. Empirical evidence demonstrates that the proposed methodology is an effective and efficient solution to prediction for data streams.
Cost%2DConstrained Data Acquisition for Intelligent Data Preparation.

Video Data Mining Semantic Indexing and Event Detection from the Association Perspective %2D Appendices.

Video Data Mining Semantic Indexing and Event Detection from the Association Perspective.
Abstract—Advances in the media and entertainment industries, including streaming audio and digital TV, present new challenges for managing and accessing large audio-visual collections. Current content management systems support retrieval using low-level features, such as motion, color, and texture. However, low-level features often have little meaning for naive users, who much prefer to identify content using high-level semantics or concepts. This creates a gap between systems and their users that must be bridged for these systems to be used effectively. To this end, in this paper, we first present a knowledge-based video indexing and content management framework for domain specific videos (using basketball video as an example). We will provide a solution to explore video knowledge by mining associations from video data. The explicit definitions and evaluation measures (e.g., temporal support and confidence) for video associations are proposed by integrating the inherent feature of video data. Our approach uses video processing techniques to find visual and audio cues (e.g., court field, camera motion activities, and applause), introduces multilevel sequential association mining to explore associations among the audio and visual cues, classifies the associations by assigning each of them with a class label, and uses their appearances in the video to construct video indices. Our experimental results demonstrate the performance of the proposed approach. Index Terms—Video mining, multimedia systems, database management, knowledge-based systems. æ 1
InsightVideo toward hierarchical video content organization for efficient browsing
Hierarchical video browsing and feature-based video retrieval are two standard methods for accessing video content. Very little research, however, has addressed the benefits of integrating these two methods for more effective and efficient video content access.
Error Detection and Impact%2DSensitive Instance Ranking in Noisy Datasets.
Given a noisy dataset, how to locate erroneous instances and attributes and rank suspicious instances based on their impacts on the system performance is an interesting and important research issue. We provide in this paper an Error Detection and Impact-sensitive instance Ranking (EDIR) mechanism to address this problem. Given a noisy dataset D, we first train a benchmark classifier T from D. The instances, that cannot be effectively classified by T are treated as suspicious and forwarded to a subset S. For each attribute A i, we switch A i and the class label C to train a classifier AP i for A i. Given an instance I k in S, we use AP i and the benchmark classifier T to locate the erroneous value of each attribute A i. To quantitatively rank instances in S, we define an impact measure based on the Information-gain Ratio (IR). We calculate IR i between attribute A i and C, and use IR i as the impact-sensitive weight of A i. The sum of impact-sensitive weights from all located erroneous attributes of I k indicates its total impact value. The experimental results demonstrate the effectiveness of our strategies. 1.

Cost%2DGuided Class Noise Handling for Effective Cost%2DSensitive Learning.
Recent research in machine learning, data mining and related areas has produced a wide variety of algorithms for costsensitive (CS) classification, where instead of maximizing the classification accuracy, minimizing the misclassification cost becomes the objective. However, these methods assume that training sets do not contain significant noise, which is rarely the case in real-world environments. In this paper, we systematically study the impacts of class noise on CS learning, and propose a cost-guided class noise handling algorithm to identify noise for effective CS learning. We call it Cost-guided Iterative Classification Filter (CICF), because it seamlessly integrates costs and an existing Classification Filter [1] for noise identification. Instead of putting equal weights to handle noise in all classes in existing efforts, CICF puts more emphasis on expensive classes, which makes it especially successful in dealing with datasets with a large cost-ratio. Experimental results and comparative studies from real-world datasets indicate that the existence of noise may seriously corrupt the performance of CS classifiers, and by adopting the proposed CICF algorithm, we can significantly reduce the misclassification cost of a CS classifier in noisy environments. 1.

Data Acquisition with Active and Impact%2DSensitive Instance Selection.

OIDM Online Interactive Data Mining.

Dealing with Predictive%2Dbut%2DUnpredictable Attributes in Noisy Data Sources.
Abstract. Attribute noise can affect classification learning. Previous work in handling attribute noise has focused on those predictable attributes that can be predicted by the class and other attributes. However, attributes can often be predictive but unpredictable. Being predictive, they are essential to classification learning and it is important to handle their noise. Being unpredictable, they require strategies different from those of predictable attributes. This paper presents a study on identifying, cleansing and measuring noise for predictive-but-unpredictable attributes. New strategies are accordingly proposed. Both theoretical analysis and empirical evidence suggest that these strategies are more effective and more efficient than previous alternatives. 1
Class Noise vs. Attribute Noise A Quantitative Study.

ClassView hierarchical video shot classification%2C indexing%2C and accessing.
Recent advances in digital video compression and networks have made video more accessible than ever. However, the existing content-based video retrieval systems still suffer from the following problems. 1 ) Semantics---sensitive video classification problem because of the semantic gap between low-level visual features and high-level semantic visual concepts; 2) Integrated video access problem because of the lack of efficient video database indexing, automatic video annotation, and concept-oriented summary organization techniques. In this paper, we have proposed a novel framework, called ClassView, to make some advances toward more efficient video database indexing and access. 1) A hierarchical semantics-sensitive video classifier is proposed to shorten the semantic gap. The hierarchical tree structure of the semantics-sensitive video classifier is derived from the domain-dependent concept hierarchy of video contents in a database. Relevance analysis is used for selecting the discriminating visual features with suitable importances. The Expectation-Maximization (EM) algorithm is also used to determine the classification rule for each visual concept node in the classifier. 2) A hierarchical video database indexing and summary presentation technique is proposed to support more effective video access over a large-scale database. The hierarchical tree structure of our video database indexing scheme is determined by the domain-dependent concept hierarchy which is also used for video classification. The presentation of visual summary is also integrated with the inherent hierarchical video database indexing tree structure. Integrating video access with efficient database indexing tree structure has provided great opportunity for supporting more powerful video search engines.
VDBMS A testbed facility for research in video database benchmarking.
Real-world video-based applications require database technology that is capable of storing digital video in the form of video databases and providing content-based video search and retrieval. Methods for handling traditional data storage, query, search, retrieval, and presentation cannot be extended to provide this functionality. The VDBMS research initiative is motivated by the requirements of videobased applications to search and retrieve portions of video data based on content and by the need for testbed facilities to facilitate research in the area of video database management. In this paper we describe the VDBMS video database research platform, a system that supports comprehensive and efficient database management for digital video. Our fundamental concept is to provide a full range of functionality for video as a well-defined abstract database data type, with its own description, parameters, and applicable methods. Research problems that are addressed by VDBMS to support the handling of video data include MPEG7 standard multimedia content representation, algorithms for image-based shot detection, image processing techniques for extracting lowlevel visual features, a high-dimensional indexing technique to access the high-dimensional feature vectors extracted by image preprocessing, multimedia query processing and optimization, new query operators, real-time stream management, a search-based buffer management policy, and an access control model for selective, content-based access to streaming video. VDBMS also provides an environment for testing the correctness and scope of new video processing techniques, measuring the performance of algorithms in a standardized way, and comparing the performance of different implementations of an algorithm or component. We are ...
Eliminating Class Noise in Large Datasets.
This paper presents a new approach for identifying and eliminating mislabeled instances in large or distributed datasets. We first partition a dataset into subsets, each of which is small enough to be processed by an induction algorithm at one time. We construct good rules from each subset, and use the good rules to evaluate the whole dataset. For a given instance Ik, two error count variables are used to count the number of times it has been identified as noise by all subsets. The instance with higher error values will have a higher probability of being a mislabeled example. Two threshold schemes, majority and non-objection, are used to identify the noise. Experimental results and comparative studies from real-world datasets are reported to evaluate the effectiveness and efficiency of the proposed approach. 1.
Mining Video Associations for Efficient Database Management.
To support more efficient video database management, this paper explores the concept of video association mining, with which the association patterns are characterized by sequentially associated video shots and their cluster information. Given a continuous video sequence V, the video shot segmentation mechanism is first introduced to parse it into discrete shots. We then cluster shots into visually distinct groups and construct a shot cluster sequence by using the class label of each shot. An association mining scheme is designed to mine sequentially associated clusters from the sequence. Those detected associations will convey valuable knowledge for video database management. The experimental results demonstrate the effectiveness of our design. 1
A hierarchical access control model for video database systems.
... In this paper, we propose a novel approach to support multilevel access control in video databases. Our access control technique combines a video database indexing mechanism with a hierarchical organization of visual concepts (i.e., video database indexing units), so that different classes of users can access different video elements or even the same video element with different quality levels according to their permissions. These video elements, which, in our access control mechanism, are used for specifying the authorization objects, can be a semantic cluster, a subcluster, a video scene, a video shot, a video frame, or even a salient object (i.e., region of interest). In the paper, we first introduce our techniques for obtaining these multilevel 
Hierarchical video content description and summarization using unified semantic and visual similarity.
Vide is increasingly the medium of choice for a variety of communication channeca renneca primarily from incre8kw le vex ofneh orke multimeST systeme One way to keW ourheSw above the vide se is to provide summarie in a more tractable format. Manye xistingapproache are limite to e xploring important low-le ve lfe9hxL re9hxL units for summarization. Unfortunatew ,the sew8WSTPk conteW andstructure of the vide do not corre9STP to low-le ve lfeNNwwS direwwS ,e ve n with close99STPLxPkS sceS deSe9STP and audio signal proceo -S8k The drawbacks ofe xisting meting are the following: (1) insteL of unfolding seolding and structure withinthe videi low-le ve l units usually addrey onlythe deS8h8 and (2) any important unitsetShhxN strate gybase on low-le ve lfe9 ture cannotbe applie togehWW8 videW8 Providingusen with an ove99w w ofthe vide contex at variousle ve8 of summarization is eSkL8WST formore e ficieP database reaba val and browsing. In thispape , we preWPW ahiexWST89xk vide conteW deteWST89 and summarizationstrate gysupporte by a nove l joint sentS9P and visual similaritystrate gy. TodeP9WkS the vide conteW e ficieWkS and accurateT , avide conteP de scription ontology isadopteL Variousvide procehxLw teoc nique are the utilize to construct asewx9ST8hPkwS vide annotationframe work. Byinte grating acquire contee de scription data, a hieNLSThxhW vide conteL structure is constructe with groupme ging and clusteWSTh Finally, a four laye vide summary with diffeSh9 granularitie isasseWPkS to assistuses in unfoldingthe vide contei in aprogreS sive way.ExpeSThW9w onrehWL ordvidek have validate the effectiveness of the proposed approach.
Accessing Video Contents through Key Objects over IP.

ClassMiner mining medical video for scalable skimming and summarization.
this paper, we define the &#034;Clinical Operation&#034; as a group of shots without speaker change, where at east one shot in the scene contains blood-red or a close-up of a skin region (skin region with size larger than 20% of the total frame size) or where more than half of the shots in the scene contain skin regions
ClassMiner Mining Medical Video Content Structure and Events Towards Efficient Access and Scalable Skimming.

A Distributed Database Server for Continuous Media.

Towards facial feature extraction and verification for omni%2Dface detection in video%2Fimages.

Qualitative Camera Motion Classification for Content%2DBased Video Indexing.

Semi%2Dautomatic Video Content Annotation.

A Video Database Management System for Advancing Video Database Research.
The most useful environments for advancing research and development in video  databases are those that provide complete video database management, including (1)  video preprocessing for content representation and indexing, (2) storage management for  video, metadata and indices, (3) image and semantic-based query processing, (4) realtime  buffer management, and (5) continuous media streaming. Such environments  support the entire process of investigating, implementing, analyzing and evaluating new  techniques, thus identifying in a concrete way which techniques are truly practical and  robust. In this paper we present a video database research initiative that culminated in the  successful development of VDBMS, a video database research platform that supports  comprehensive and efficient database management for digital video. We describe key  video processing components of the system and illustrate the value of VDBMS as a  research platform by describing several research projects carried out within the VDBMS  environment. These include MPEG7 document support for video feature import and  export, a new query operator for optimal multi-feature image similarity matching, secure  access control for streaming video, and the mining of medical video data using  hierarchical content organization.
Seeded image segmentation for content%2Dbased image retrieval application.

Hierarchical video summarization for medical data.
To provide users with an overview of medical video content at various levels of abstraction which can be used for more efficient database browsing and access, a hierarchical video summarization strategy has been developed and is presented in this paper. To generate an overview, the key frames of a video are preprocessed to extract special frames (black frames, slides, clip art, sketch drawings) and special regions (faces, skin or blood-red areas). A shot grouping method is then applied to merge the spatially or temporally related shots into groups. The visual features and knowledge from the video shots are integrated to assign the groups into predefined semantic categories. Based on the video groups and their semantic categories, video summaries for different levels are constructed by group merging, hierarchical group clustering and semantic category selection. Based on this strategy, a user can select the layer of the summary to access. The higher the layer, the more concise the video summary; the lower the layer, the greater the detail contained in the summary.
Model%2DBased Video Classification toward Hierarchical Representation

Seeded Semantic Object Generation toward Content%2DBased Video Indexing.

Automatic Scene Detection in News Program by Integrating Visual Feature and Rules.

Automatic model%2Dbased semantic object extraction algorithm.


New query refinement and semantics integrated image retrieval system with semiautomatic annotation scheme.
Relevance feedback is a powerful and widely used technique in content-based image retrieval systems. However, most relevance feedback approaches use only weighted feature sum of the feedback images to optimize the query for refining image similarity assessment. Such approaches do not work very well in most cases, especially when the user wants to express an &#034;OR&#034; relationship among the queries. In this paper, we propose three methods, weighted distance sum, minimal distance (MD), and minimal distance rank (MDR), to measure the similarity between images in database and the feedback images in query refinement. After experimental comparisons we propose a relevance feedback scheme using the MDR method and the MD method to describe the user&#039;s multiple intentions. Based on this scheme, an image retrieval and semiautomatic annotation system, iFind [Y. Lu, C. H. Hu, X. Q. Zhu, H. J. Zhang, and Q. Yang, &#034;A unified framework for semantics and feature based relevance feedback in image retrieval systems,&#034; in Proc. 8th ACM International Conference, pp. 31--37 (2000) and X. Q. Zhu, W. Y. Liu, H. J. Zhang, and L. D. Wu, &#034;An image retrieval and semi-automatic annotation scheme for large image databases on the web,&#034; in SPIE photonics West (Jan. 2001), Electronic Imaging 4311], which integrates query refinement and semantic information, is presented. Experiments show that the proposed methods can result in substantial improvement in retrieval accuracy and can be especially useful for retrieval or annotating large image databases.
MultiView Multilevel video content representation and retrieval.
In this article, several practical algorithms are proposed to support content-based video analysis, modeling, representation, summarization, indexing, and access. First, a multilevel video database model is given. One advantage of this model is that it provides a reasonable approach to bridging the gap between low-level representative features and high-level semantic concepts from a human point of view. Second, several model-based video analysis techniques are proposed. In order to detect the video shots, we present a novel technique, which can adapt the threshold for scene cut detection to the activities of variant videos or even different video shots. A seeded region aggregation and temporal tracking technique is proposed for generating the semantic video objects. The semantic video scenes can then be generated from these extracted video access units (e.g., shots and objects) according to some domain knowledge. Third, in order to categorize video contents into a set of semantic clusters, an integrated video classification technique is developed to support more efficient multilevel video representation, summarization, indexing, and access techniques. 2001 SPIE and IS&amp;T. [DOI: 10.1117/1.1406944]  1 

