Improved coding of 3D spatio-temporal wavelet video is achieved using matching pursuits (MP) with a separable 2D basis dictionary and temporal replication of atoms. Using two temporal scales and the biorthogonal 5/3 wavelet filter incorporating motion compensation, four sets of four transform planes are combined into a group of temporal planes (GOTP). The planes are wavelet transformed to 4 scales spatially before the 2D MP algorithm is applied over the entire GOTP to select the optimum inner product with a 2D dictionary of 64 bases formed separably from 8 ID basis functions. The selected atom is conditionally replicated in the three corresponding temporal frequency planes within the GOTP. Compared to applying 2D MP to the planes without replication, a gain of up to 2 dB has been achieved. Comparisons with a standard H.263 codec and the classic MP enhanced H.263 MP codec, show significant gains at low bit rates.

3D wavelet video coding with replicated matching pursuits

L'ensemble des travaux presentes traite le cas des signaux issus des milieux naturels et/ou des milieux incertains, ce qui se traduit par des cas d'analyse complexes. Les solutions d'analyse ne peuvent demeurer que dans le vaste domaine de l'analyse des signaux non-stationnaire et c'est la raison pour laquelle trois grandes classes de methodes constituent le point de depart : l'analyse temps-frequence lineaire et bilineaire, les transformations non-lineaires (les operateurs warping) et les transformations d'ordre superieur (modelisation polynomiale de phase). A partir de ces trois piliers, la construction des methodes de caracterisation parcimonieuse des signaux a composantes temps-frequence non-lineaires represente l'objectif de base de mes travaux. Ainsi, trois axes de recherche ont ete investigues. Le premier axe a ete constitue par le tracking des structures temps-frequence arbitraires en utilisant le regroupement des atomes temps-frequence par un algorithme de type Viterbi. Le deuxieme axe a eu pour objectif la generalisation des operateurs warping pour des modulations complexes, par forcement inversibles mathematiquement. Enfin, la definition des distributions a temps complexe a constitue le troisieme axe. L'ensemble de ces axes a ete continue en cherchant a mieux analyser les signaux a structures temps-frequence non-lineaires. Ainsi, une contribution importante a ete constituee par le filtrage temps-frequence base sur le warping genealise qui permet l'extraction des structures temps-frequence non-lineaires en utilisant des operateurs de traitement mono-dimensionnels. Cette contribution nous a permis de definir la classe des methodes de suivi temps-frequence-phaseen utilisant egalement des techniques de modelisation polynomiale de phase ainsi que le principe du tracking par la methode de Viterbi. Cette classe de methode de suivi temps-frequence a prouve son caractere general ainsi que son efficacite dans des applications diverses.

https://tel.archives-ouvertes.fr/tel-00985501/document

Caratérisation des signaux non-stationnaires à composantes temps-fréquence non-linéaires

Discrimination-based classifiers differentiate between two classes by drawing a decision boundary between their data members in the feature domain. These classifiers are capable of correctly labeling the test data that belongs to the same distribution as the training data. However, since the decision boundary is meaningless beyond the training points, the class label of an outlier determined with respect to this extended decision boundary will be a random value. Therefore, discrimination-based classifiers lack a mechanism for outlier detection in the test data. To counter this problem, a prototype-based classifier may be used that assigns class label to a test point based on its similarity to the prototype of that class. If a test point is dissimilar to all class prototypes, it may be considered an outlier. 
Prototype-based classifiers are usually clustering-based methods. Therefore, they require a dissimilarity criterion to cluster the training data and also to assign class labels to test data. Euclidean distance is a commonly used dissimilarity criterion. However, the Euclidean distance may not be able to give accurate shape-based comparisons of very high-dimensional signals. This can be problematic for some classification applications where high-dimensional signals are grouped into classes based on shape similarities. Therefore, a reliable shape-based dissimilarity measure is desirable. 
In order to be able to build reliable prototype-based classifiers that can utilize shape-based information for classification, we have developed a matching pursuits dissimilarity measure (MPDM). The MPDM is capable of performing shape-based comparisons between very high-dimensional signals. The MPDM extends the matching pursuits (MP) algorithm [1] which is a well-known signal approximation method. The MPDM is a versatile measure as it can also be adopted for magnitude-based comparisons between signals, similar to the Euclidean distance. 
The MPDM has been used with the competitive agglomeration fuzzy clustering algorithm (CA) [2] to develop a prototype-based probabilistic classifier, called CAMP. The CAMP algorithm is the first method of its kind as it builds a bridge between clustering and matching pursuits algorithms. The preliminary experimental results also demonstrate its superior performance over a neural network classifier and a prototype-based classifier using the Euclidean distance. The performance of CAMP has been tested on high-dimensional synthetic data and also on real landmines detection data. 
The MPDM is also used to develop an automated dictionary learning algorithm for MP approximation of signals. This algorithm uses the MPDM and the CA clustering algorithm to learn the required number of dictionary elements during training. Under-utilized and replicated dictionary elements are gradually pruned to produce a compact dictionary, without compromising its approximation capabilities. The experimental results show that the size of the dictionary learned by our method is 60% smaller but with same approximation capabilities as the existing dictionary learning algorithms.

Optimized dictionary design and classification using the matching pursuits dissimilarity measure

The retrieval of soccer highlights is a suitable technique for video indexing, required by the multimedia database management or for the development of television on demand. For these purposes, it should be interesting to have an automatic annotation of events happened in soccer games. One solution consists in analyzing the audio soundtrack associated to the soccer video and to detect the interesting frames. In this paper we use the adaptive time-frequency decomposition of the soundtrack as a feature extraction procedure. This decomposition is based on the Matching Pursuit concept and a dictionary composed of Gabor functions. The parameters provided by these transformations constitute the input of the classification stage. The results provided for real soccer video will prove the efficiency of the adaptive time-frequency representation as a feature extraction stage.

J. Marchal

Papers

Soccer Video Retrival Using Adaptive Time-Frequency Methods