Search

This abstract states the position of the Publimape project, and unveils progress achieved since its recent start.

An important contribution of satellite Sea Surface Salinity (SSS) is the spatio-temporal monitoring of rivers fresh water plumes at mesoscale. In this paper, we detail a new correction for systematic errors in the Soil Moisture and Ocean Salinity (SMOS) measurements that is implemented in the Centre Aval de Traitement des Donnees SMOS (CATDS). With this new mitigation, the SMOS and Soil Moisture Active Passive (SMAP) SSS monitor very consistent features in most areas close to continents. The rms-difference between bi-weekly SMOS and SMAP SSS over 20 months and in selected coastal regions is about 0.3pss (once outliers are filtered out), rather consistent with the rms-difference between satellite and in situ SSS (on the order of 0.2pss). The coefficient of determination (r2) between SMOS and SMAP SSS is above than 0.8 in very fresh areas (river plumes). Over the open ocean, the rms difference between SMOS and ship SSS is 0.2pss.

We present an approach for unsupervised computation of local shape descriptors, which relies on the use of linear autoencoders for characterizing local regions of complex shapes. The proposed approach responds to the need for a robust scheme to index binary images using local descriptors, which arises when only few examples of the complete images are available for training, thus making inaccurate the learning process of parameters of traditional neural networks schemes. Given the possibility of using linear operations during the encoding phase, the computation of the proposed local descriptor can be fast once the parameters of the encoding function are learned. After conducting a vast search, we identified the optimal dimensionality of the resulting local descriptor to be of only 128 dimensions, which allows for efficient further operations on them, such as the construction of bag representations with purposes of shape retrieval and classification. We evaluated the proposed approach indexing a collection of complex binary images, whose instances contain compounds of hieroglyphs from the ancient Maya civilization. Our retrieval experiments show that the proposed approach achieves competitive retrieval performance when compared with hand-crafted local descriptors.

Permutation-based indexing is one of the most popular techniques for the approximate nearest-neighbor search problem in high-dimensional spaces. Due to the exponential increase of multimedia data, the time required to index this data has become a serious constraint of the indexing techniques. One of the possible steps towards faster index construction is utilization of massively parallel platforms such as the GPGPU architectures. In this paper, we have analyzed the computational costs of individual steps of the permutation-based index construction in a high-dimensional feature space and proposed a hybrid solution, where computational power of GPU is utilized for distance computations whilst the host CPU performs the postprocessing and sorting steps. Despite the fact that computing the distances is a naturally data-parallel task, an efficient implementation is quite challenging due to various GPU limitations and complex memory hierarchy. We have tested possible approaches to work division and data caching to utilize the GPU to its best abilities. We summarize our empirical results and point out the optimal solution.

Searching for digital images in large-scale multimedia database is a hard problem due to the rapid increase of the digital assets. Metric Permutation Table is an efficient data structure for large-scale multimedia indexing. This data structure is based on the Permutation-based indexing, that aims to predict the proximity between elements encoding their location with respect to their surrounding. The main constraint of the Metric Permutation Table is the indexing time. With the exponential increase of multimedia data, parallel computation is needed. Opening the GPUs to general purpose computation allows to perform parallel computation on a powerful platform. In this paper, we propose efficient indexing and searching algorithms for the Metric Permutation Table using GPU and multi-core CPU. We study the performance and efficiency of our algorithms on large-scale datasets of millions of images. Experimental results show a decrease of the indexing time while preserving the quality of the results.

Service Composition is an essential part of Service Oriented Architectures (SOA). The number of available alternatives for web services offering similar functional characteristics opens opportunities to maximise compositions based on criteria embedded into the Quality of Service (QoS). These non-functional criteria may be numerous and diverse, and arise from various constraints, making composition an optimal information fusion problem. A number of solutions have been offered to resolve the selection problem for web service composition, essentially relying on exploiting and combining absolute performance values for measuring multi-dimensional QoS. In this paper, we investigate the use of Rank Aggregation methods for performing adequate information fusion for web service selection in view of composition. Our goal is to take advantage of these methods working on ranks rather than on the direct values for inferring global QoS. We formalise our approach and demonstrate its validity via an empirical study. We further evaluate our approach on publicly available real data gathered from actual web service characteristics.

This paper extends the bag-of-visual-words representations to a bag-of-visual-phrases model. The introduced bag-of-visual-phrases representation is constructed upon a proposed method for probabilistic description of co-occurring visual words, which is adapted for each reference word. This bag-of-visual-phrases representation implicitly encodes spatial relationships among visual words, thus being a richer representation while remaining as compact as the bag-of-visual-words model. We demonstrate the effectiveness of our method with a series of statistical analysis and retrieval experiments, and show that it largely outperforms previous methods for construction of bag representations. Furthermore, our method allows to query traditional bag-of-words vs the proposed bag-of-phrases. We conducted retrieval experiments on a dataset of complex shapes, whose instances correspond to hieroglyphs of the pre-Columbian Maya culture from the ancient Americas.

Massive amounts of textual and digital data are created daily from business or public activities. The organisation, mining and summarization of such a rich and large information source is required to capture the essential and critical knowledge it contains. Such a mining is of strategic importance in many domains including innovation (eg to mine technological reviews and scientific literature) and electronic commerce (eg to mine customer reviews). Information content generally bears several important aspects, mapped onto visualisation dimensions, whose number needs to be reduced to enable relevant interactive exploration. In this paper, we propose a novel strategy to mine and organise document sets, in order to present them in a consistent manner and to highlight interesting and relevant information patterns they contain. We base our method on the formulation of a global optimisation problem solved by using the Traveling Salesman Problem (TSP) approach. We show how this compact formulation opens interesting possibilities for the mining of document collections mapped onto multidimensional information sets. We discuss the issue of scalability and show that associated scalable solutions exist. We demonstrate the effectiveness of our method over several types of documents, embedded into real business cases.

Web-scale digital assets comprise millions or billions of documents. Due to such increase, sequential algorithms cannot cope with this data, and parallel and distributed computing become the solution of choice. MapReduce is a programming model proposed by Google for scalable data processing. MapReduce is mainly applicable for data intensive algorithms. In contrast, The message passing interface (MPI) is suitable for high performance algorithms. This paper proposes an adapted structure of MapReduce programming model using MPI for multimedia indexing. Experimental results on a large number of text (XML) excerpts related to images from the ImageNet corpus indicate that our implementation achieved good speedup compared to the sequential version and the earlier versions of MapReduce using MPI. Extensions to index large-scale multimedia collections are discussed.

MapReduce is a programming model proposed by Google to simplify large-scale data processing. In contrast, the message passing interface (MPI) standard is extensively used for algorithmic parallelization, as it accommodates an efficient communication infrastructure. In the original implementation of MapReduce, the reduce function can only start processing following termination of the map function. If the map function is slow for any reason, this will affect the whole running time. In this paper, we propose MapReduce overlapping using MPI, which is an adapted structure of the MapReduce programming model for fast intensive data processing. Our implementation is based on running the map and the reduce functions concurrently in parallel by exchanging partial intermediate data between them in a pipeline fashion using MPI. At the same time, we maintain the usability and the simplicity of MapReduce. Experimental results based on two different applications (Word Count and Distributed Inverted Indexing) show a good speedup compared to the earlier versions of MapReduce such as Hadoop and the available MPI-MapReduce implementations. For word count, we are able to achieve 1.9x and 5.3x speedup comparing to Hadoop and MPI-MapReduce respectively for 53Gb of data.

The multiplication of large-scale document collections has created the need for robust and adaptive access strategies in many applicative areas. In this paper, we depart from the traditional document search paradigm to move onto the construction of a collection navigation strategy. We thus detail a model where user clicks are taken as expression of interest rather than positive search feedback. In terms of the interpretation of user interaction, our model is close to the recommendation paradigm. However, it exploits user feedback in a navigational procedure to perform an approximation of interactive metric learning procedures. Our model is directly compatible with distributed data indexing and is therefore inherently built to cope with scalability issues. The evaluation of such models is discussed and implemented over the context of an image collection. Experiments are presented where the user browses a collection of images with changing center of focus in the course of the browsing operations. Results show that our model is effectively able to recommend the user a relevant path through the collection and to adapt according to variable expressed interests.

In this paper we formalise a query simulation framework for the evaluation of long-term learning systems for image retrieval. Long-term learning relies on historical queries and associated relevance judgements, usually stored in query logs, in order to improve search results presented to users of the retrieval system. Evaluation of long-term learning methods requires access to query logs, preferably in large quantity. However, real-world query logs are notoriously difficult to acquire due to legitimate efforts of safeguarding user privacy. Query log simulation provides a useful means of evaluating long-term learning approaches without the need for real-world data. We introduce a query log simulator that is based on a user model of long-term learning that explains the observed relevance judgements contained in query logs. We validate simulated queries against a real-world query log of an image retrieval system and demonstrate that for evaluation purposes, the simulator is accurate on a global level.

Multimedia data collections immersed into social networks may be explored from the point of view of varying documents and users characteristics. In this paper, we develop a unified model to embed documents and users into coherent structures from which to extract optimal subsets. The result is the definition of guiding navigation strategies of both the user and document networks, as a complement to classical search operations. An initial interface that may materialize such browsing over documents is demonstrated in the context of Cultural Heritage.

Information fusion, especially for high dimensional multimedia data, is still an open research problem. In this article, we present a new approach to target this problem. Feature information interaction is an information-theoretic dependence measure that can determine synergy and redundancy between attributes, which then can be exploited with feature selection and construction towards more efficient information fusion. This also leads to improved performances for algorithms that rely on information fusion like multimedia document classification. We show that synergetic and redundant feature pairs require different fusion strategies for optimal exploitation. The approach is compared to classical feature selection strategies based on correlation and mutual information.

User-supplied data such as browsing logs, click-through data, and relevance feedback judgements are an important source of knowledge during semantic indexing of documents such as images and video. Low-level indexing and abstraction methods are limited in the manner with which semantic data can be dealt. In this paper and in the context of this semantic data, we apply latent semantic analysis on two forms of user-supplied data, real-world and artificially generated relevance feedback judgements in order to examine the validity of using artificially generated interaction data for the study of semantic image clustering.

This work concerns the problem of asymmetric classification and provides the following contributions. First, it introduces the method of KDDA - a kernelized extension of the distance-based discriminant analysis technique that treats data asymmetrically and naturally accommodates indefinite kernels. Second, it demonstrates that KDDA and other asymmetric nonlinear projective approaches, such as BiasMap and KFD are often prone to an adverse condition referred to as the false positive projection effect. Empirical evaluation on both synthetic and real-world data sets is carried out to assess the degree of performance degradation due to false positive projection effect, determine the viability of some schemes for its elimination, and compare the introduced KDDA method with state-of-the-art alternatives, achieving encouraging results

This paper presents a hierarchical ensemble learning method applied in the context of multimedia autoannotation. In contrast to the standard multiple-category classification setting that assumes independent, non-overlapping and exhaustive set of categories, the proposed approach models explicitly the hierarchical relationships among target classes and estimates their relevance to a query as a trade-off between the goodness of fit to a given category description and its inherent uncertainty. The promising results of the empirical evaluation confirm the viability of the proposed approach, validated in comparison to several techniques of ensemble learning, as well as with different type of baseline classifiers

This paper formulates the problem of object categorization in the discriminant analysis framework focusing on transforming visual feature data so as to make it conform to the compactness hypothesis in order to improve categorization accuracy. The sought transformation, in turn, is found as a solution to an optimization problem formulated in terms of inter-observation distances only, using the technique of iterative majorization. The proposed approach is suitable for both binary and multiple-class categorization problems, and can be applied as a dimensionality reduction technique. In the latter case, the number of discriminative features is determined automatically since the process of feature extraction is fully embedded in the optimization procedure. Performance tests validate our method on a number of benchmark data sets from the UCI repository, while the experiments in the application of visual object and content-based image categorization demonstrate very competitive results, asserting the method's capability of producing semantically relevant matches that share the same or synonymous vocabulary with the query category and allowing multiple pertinent category assignment.

While one-dimensional hidden Markov models have been very successfully applied to numerous problems, their extension to two dimensions has been shown to be exponentially complex, and this has very much restricted their usage for problems such as image analysis. In this paper we propose a novel algorithm which is able to approximate the search for the best state path (Viterbi decoding) in a 2D HMM. This algorithm makes certain assumptions which lead to tractable computations, at a price of loss in full optimality. We detail our algorithm, its implementation, and present some experiments on handwritten character recognition. Because the Viterbi algorithm serves as a basis for many applications, and 1D HMMs have shown great flexibility in their usage, our approach has the potential to make 2D HMMs as useful for 2D data as 1D HMMs are for 1D data such as speech.