Atilla Baskurt

Sequential deep learning for human action recognition

We propose in this paper a fully automated deep model, which learns to classify human actions without using any prior knowledge. The first step of our scheme, based on the extension of Convolutional Neural Networks to 3D, automatically learns spatio-temporal features. A Recurrent Neural Network is then trained to classify each sequence considering the temporal evolution of the learned features for each timestep. Experimental results on the KTH dataset show that the proposed approach outperforms existing deep models, and gives comparable results with the best related works.

A new CAD mesh segmentation method, based on curvature tensor analysis

This paper presents a new and efficient algorithm for the decomposition of 3D arbitrary triangle meshes and particularly optimized triangulated CAD meshes. The algorithm is based on the curvature tensor field analysis and presents two distinct complementary steps: a region based segmentation, which is an improvement of that presented by Lavoue et al. [Lavoue G, Dupont F, Baskurt A. Constant curvature region decomposition of 3D-meshes by a mixed approach vertex-triangle, J WSCG 2004;12(2):245-52] and which decomposes the object into near constant curvature patches, and a boundary rectification based on curvature tensor directions, which corrects boundaries by suppressing their artefacts or discontinuities. Experiments conducted on various models including both CAD and natural objects, show satisfactory results. Resulting segmented patches, by virtue of their properties (homogeneous curvature, clean boundaries) are particularly adapted to computer graphics tasks like parametric or subdivision surface fitting in an adaptive compression objective.

Segmentation of ultrasound images: multiresolution 2D and 3D algorithm based on global and local statistics

In this paper, we propose a robust adaptive region segmentation algorithm for noisy images, within a Bayesian framework. A multiresolution implementation of the algorithm is performed using a wavelets basis and can be used to process both 2D and 3D data. In this work we focus on the adaptive character of the algorithm and we discuss how global and local statistics can be utilised in the segmentation process. We propose an improvement on the adaptivity by introducing an enhancement to control the adaptive properties of the segmentation process. This takes the form of a weighting function accounting for both local and global statistics, and is introduced in the minimisation. A new formulation of the segmentation problem allows us to control the effective contribution of each statistical component. The segmentation algorithm is demonstrated on synthetic data, 2D breast ultrasound data and on echocardiographic sequences ð2D þ TÞ. An evaluation of the performance of the proposed algorithm is also presented. � 2002 Elsevier Science B.V. All rights reserved.

A Comprehensive Survey on Three-Dimensional Mesh Watermarking

Three-dimensional (3-D) meshes have been used more and more in industrial, medical and entertainment applications during the last decade. Many researchers, from both the academic and the industrial sectors, have become aware of their intellectual property protection and authentication problems arising with their increasing use. This paper gives a comprehensive survey on 3-D mesh watermarking, which is considered an effective solution to the above two emerging problems. Our survey covers an introduction to the relevant state of the art, an attack-centric investigation, and a list of existing problems and potential solutions. First, the particular difficulties encountered while applying watermarking on 3-D meshes are discussed. Then we give a presentation and an analysis of the existing algorithms by distinguishing them between fragile techniques and robust techniques. Since attacks play an important role in the design of 3-D mesh watermarking algorithms, we also provide an attack-centric viewpoint of this state of the art. Finally, some future working directions are pointed out especially on the ways of devising robust and blind algorithms and on some new probably promising watermarking feature spaces.

Generalizations of angular radial transform for 2D and 3D shape retrieval

The angular radial transform (ART) is a moment-based image description method adopted in MPEG-7 as a 2D region-based shape descriptor. This paper proposes generalizations of the ART to describe two-dimensional images and three-dimensional models. First, we propose an 2D extension, called GART, which allows applying ART to images while insuring robustness to all possible rotations and to perspective deformations. Then, we generalize the ART to index 3D models. This new 3D shape descriptor, so called 3D ART, has the same properties that the original transform: robustness to rotation, translation, noise and scaling while keeping a compact size and a good retrieval cost. The size of the descriptor is an essential evaluation parameter on which depends the response time of a content-based retrieval system. For both generalizations, many experiments were made on large databases and have shown, that GART outperforms ART in accuracy at the cost of speed, and that 3D ART outperforms the spherical harmonics shape descriptor (Vranic, D.V., Saupe, D., 2002. Description of 3D-shape using a complex function on the sphere, in: IEEE International Conference on Multimedia and Expo (ICME 2002), Lausanne, Switzerland, 2002, pp. 177-180; Funkhouser, T., Min, P. Kazhdan, M., Chen, J., Halderman, A., Dobkin, D., Jacobs, D., 2003. A search engine for 3D models. ACM Trans. Graphics 22(1), 83-105) in speed at the cost of accuracy.

Improving Zernike Moments Comparison for Optimal Similarity and Rotation Angle Retrieval

Zernike moments constitute a powerful shape descriptor in terms of robustness and description capability. However the classical way of comparing two Zernike descriptors only takes into account the magnitude of the moments and loses the phase information. The novelty of our approach is to take advantage of the phase information in the comparison process while still preserving the invariance to rotation. This new Zernike comparator provides a more accurate similarity measure together with the optimal rotation angle between the patterns, while keeping the same complexity as the classical approach. This angle information is particularly of interest for many applications, including 3D scene understanding through images. Experiments demonstrate that our comparator outperforms the classical one in terms of similarity measure. In particular the robustness of the retrieval against noise and geometric deformation is greatly improved. Moreover, the rotation angle estimation is also more accurate than state-of-the-art algorithms.

Technical Section: Robust and blind mesh watermarking based on volume moments

This paper presents a robust and blind watermarking algorithm for three-dimensional (3D) meshes. The watermarking primitive is an intrinsic 3D shape descriptor: the analytic and continuous geometric volume moment. During watermark embedding, the input mesh is first normalized to a canonical and robust spatial pose by using its global volume moments. Then, the normalized mesh is decomposed into patches and the watermark is embedded through a modified scalar Costa quantization of the zero-order volume moments of some selected candidate patches. Experimental results and comparisons with the state of the art demonstrate the effectiveness of the proposed approach.

A multiparametric and multiresolution segmentation algorithm of 3D ultrasonic data

An algorithm devoted to the segmentation of 3-D ultrasonic data is proposed. The algorithm involves 3-D adaptive clustering based on multiparametric information: the gray-scale intensity of the echographic data, 3-D texture features calculated from the envelope data, and 3-D tissue characterization information calculated from the local frequency spectra of the radio-frequency signals. The segmentation problem is formulated as a maximum a posterior (MAP) estimation problem. A multi-resolution implementation of the algorithm is proposed. The approach is tested on simulated data and on in vivo echocardiographic 3-D data. The results presented in the paper illustrate the robustness and the accuracy of the proposed approach for the segmentation of ultrasonic data.

The FG 2015 Kinship Verification in the Wild Evaluation

The aim of the Kinship Verification in the Wild Evaluation (held in conjunction with the 2015 IEEE International Conference on Automatic Face and Gesture Recognition, Ljubljana, Slovenia) was to evaluate different kinship verification algorithms. For this task, two datasets were made available and three possible experimental protocols (unsupervised, image-restricted, and image-unrestricted) were designed. Five institutions submitted their results to the evaluation: (i) Politecnico di Torino, Italy; (ii) LIRIS-University of Lyon, France; (iii) Universidad de Las Palmas de Gran Canaria, Spain; (iv) Nanjing University of Aeronautics and Astronautics, China; and (v) Bar Ilan University, Israel. Most of the participants tackled the image-restricted challenge and experimental results demonstrated better kinship verification performance than the baseline methods provided by the organizers.

Spatio-Temporal Convolutional Sparse Auto-Encoder for Sequence Classification.

We present in this paper a novel learning-based approach for video sequence classification. Contrary to the dominant methodology, which relies on hand-crafted features that are manually engineered to be optimal for a specific task, our neural model automatically learns a sparse shift-invariant representation of the local 2D+t salient information, without any use of prior knowledge. To that aim, a spatio-temporal convolutional sparse auto-encoder is trained to project a given input in a feature space, and to reconstruct it from its projection coordinates. Learning is performed in an unsupervised manner by minimizing a global parametrized objective function. The sparsity is ensured by adding a sparsifying logistic between the encoder and the decoder, while the shift-invariance is handled by including an additional hidden variable to the objective function. The temporal evolution of the obtained sparse features is learned by a long short-term memory recurrent neural network trained to classify each sequence. We show that, since the feature learning process is problem-independent, the model achieves outstanding performances when applied to two different problems, namely human action and facial expression recognition. Obtained results are superior to the state of the art on the GEMEP-FERA dataset and among the very best on the KTH dataset.