Carl Frélicot

Content-Based Copy Retrieval Using Distortion-Based Probabilistic Similarity Search

Content-based copy retrieval (CBCR) aims at retrieving in a database all the modified versions or the previous versions of a given candidate object. In this paper, we present a copy-retrieval scheme based on local features that can deal with very large databases both in terms of quality and speed. We first propose a new approximate similarity search technique in which the probabilistic selection of the feature space regions is not based on the distribution in the database but on the distribution of the features distortion. Since our CBCR framework is based on local features, the approximation can be strong and reduce drastically the amount of data to explore. Furthermore, we show how the discrimination of the global retrieval can be enhanced during its post-processing step, by considering only the geometrically consistent matches. This framework is applied to robust video copy retrieval and extensive experiments are presented to study the interactions between the approximate search and the retrieval efficiency. Largest used database contains more than 1 billion local features corresponding to 30000 h of video

Robust content-based video copy identification in a large reference database

This paper proposes a novel scheme for video content-based copy identification dedicated to TV broadcast with a reference video database exceeding 1000 hours of video. It enables the monitoring of a TV channel in soft real-time with a good tolerance to strong transformations that one can meet in any TV post-production process like: clipping, cropping, shifting, resizing, objects encrusting or color variations. Contrary to most of the existing schemes, the recognition is not based on global features but on local features extracted around interest points. This allows the selection and the localization of fully discriminant local patterns which can be compared according to a distance measure. Retrieval is performed using an efficient approximate Nearest Neighbors search and a final decision based on several matches cumulated in time.

A Cluster-Validity Index Combining an Overlap Measure and a Separation Measure Based on Fuzzy-Aggregation Operators

Since a clustering algorithm can produce as many partitions as desired, one needs to assess their quality in order to select the partition that most represents the structure in the data, if there is any. This is the rationale for the cluster-validity (CV) problem and indices. This paper presents a CV index that helps to find the optimal number of clusters of data from partitions generated by a fuzzy-clustering algorithm, such as the fuzzy c-means (FCM) or its derivatives. Given a fuzzy partition, this new index uses a measure of multiple cluster overlap and a separation measure for each data point, both based on an aggregation operation of membership degrees. Experimental results on artificial and benchmark datasets are given to demonstrate the performance of the proposed index, as compared with traditional and recent indices.

An eXtended Center-Symmetric Local Binary Pattern for Background Modeling and Subtraction in Videos

In this paper, we propose an eXtended Center-Symmetric Local Binary Pattern (XCS-LBP) descriptor for background modeling and subtraction in videos. By combining the strengths of the original LBP and the similar CS ones, it appears to be robust to illumination changes and noise, and produces short histograms, too. The experiments conducted on both synthetic and real videos (from the Background Models Challenge) of outdoor urban scenes under various conditions show that the proposed XCS-LBP outperforms its direct competitors for the background subtraction task.

Content-based video copy detection in large databases: a local fingerprints statistical similarity search approach

Recent methods based on interest points and local fingerprints have been proposed to perform robust CBCD (content-based copy detection) of images and video. They include two steps: the search for similar local fingerprints in the database (DB) and a voting strategy that merges all the local results in order to perform a global decision. In most image or video retrieval systems, the search for similar features in the DB is performed by a geometrical query in a multidimensional index structure. Recently, the paradigm of approximate k-nearest neighbors query has shown that trading quality for time can be widely profitable in that context. In this paper, we evaluate a new approximate search paradigm, called statistical similarity search (S/sup 3/) in a complete CBCD scheme based on video local fingerprints. Experimental results show that these statistical queries allow high performance gains compared to classical e-range queries and that trading quality for time during the search does not degrade seriously the global robustness of the system, even with very large DBs including more than 20,000 hours of video.

Statistical similarity search applied to content-based video copy detection

Content-based copy detection (CBCD) is one of the emerging multimedia applications for which there is a need of a concerted effort from the database community and the computer vision community. Recent methods based on interest points and local fingerprints have been proposed to perform robust CBCD of images and video. They include two steps: the search of similar fingerprints in the database and a voting strategy that merges all the local results in order to perform a global decision. In most image or video retrieval systems, the search of similar features in the database is performed by a geometrical query in a multidimensional index structure. Recently, the paradigm of approximate knearest neighbors query has shown that trading quality for time can be widely profitable in that context. In this paper, we introduce a new approximate search paradigm, called Statistical Similarity Search (S3), dedicated to local fingerprints and we describe the original indexing structure we have developped to compute efficiently the corresponding queries. The key-point relates to the distribution of the relevant fingerprints around a query. Since a video query can result from (a combination of ) more or less transformations of an original one, we modelize the distribution of the distorsion vector between a referenced fingerprint and a candidate one. Experimental results show that these statistical queries allow high performance gains compared to classical e-range queries. By studying the influence of this approximate search on a complete CBCD scheme based on local video fingerprints, we also show that trading quality for time during the search does not degrade seriously the global robustness of the system, even with very large databases including more than 20,000 hours of video.

Feature statistical retrieval applied to content based copy identification

In many image or video retrieval systems, the search for similar objects in the database includes a spatial access method to a multidimensional feature space. This step is generally considered as a problem independent of the features and the similarity type. The well known multidimensional nearest neighbor search has also been widely studied by the database community as a generic method. We propose a novel strategy dedicated to pseudo-invariant features retrieval and more specifically applied to content based copy identification. The range of a query is computed during the search according to deviation statistics between original and observed features. Furthermore, this approximate search range is directly mapped onto a Hilbert space-filling curve, allowing an efficient access to the database. Experimental results give excellent response times for very large databases both on synthetic and real data. This work is used in a TV monitoring system including more than 13000 hours of video in the reference database.

Automatic analysis of the structuring of children's drawings and writing

The aim of this work was to build an objective tool for the detection of graphomotor difficulties involving disorders in the writing of children. We outline some characteristics of layouts, describing the automation level of the graphic activity. We have defined exercises, like copying figures or writing sentences under different conditions that allowed us to measure simple aspects of graphomotor skill up to complex ones. A tool was conceived which was able to automatically extract low-level and high-level primitives. Based on such descriptors, we focus on the analysis of the temporal structuring of two particular drawings. In the final part, we present the method we used to select features that can describe the automation level of the graphic activity and we show that, in most cases, these features allow to discriminate children with academic difficulties.

A fast fuzzy c-means algorithm for color image segmentation

Color image segmentation is a fundamental task in many computer vision problems. A common approach is to use fuzzy iterative clustering algorithms that provide a partition of the pixels into a given number of clusters. However, most of these algorithms present several drawbacks: they are time consuming, and sensitive to initialization and noise. In this paper, we propose a new fuzzy c-means algorithm aiming at correcting such drawbacks. It relies on a new efficient cluster centers initialization and color quantization allowing faster and more accurate convergence such that it is suitable to segment very large color images. Thanks to color quantization and a new spatial regularization, the proposed algorithm is also more robust. Experiments on real images show the efficiency in terms of both accuracy and computation time of the proposed algorithm as compared to recent methods of the literature.

A k-order fuzzy OR operator for pattern classification with k -order ambiguity rejection

In pattern recognition, the membership of an object to classes is often measured by labels. This article mainly deals with the mathematical foundations of labels combination operators, built on t-norms, that extend previous ambiguity measures of objects by dealing not only with two classes ambiguities but also with k classes, k lying between 1 and the number of classes c. Mathematical properties of this family of combination operators are established and a weighted extension is proposed, allowing to give more or less importance to a given class. A classifier with reject options built on the proposed measure is presented and applied on synthetic data. A critical analysis of the results led to derivate some new operators by aggregating previous measures. A modified classifier is proposed and applied to synthetic data as well as to standard real data.