Francesco Pelosin, Saurav Jha, Andrea Torsello, Bogdan Raducanu, & Joost Van de Weijer. (2022). Towards exemplar-free continual learning in vision transformers: an account of attention, functional and weight regularization. In IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW).
Abstract: In this paper, we investigate the continual learning of Vision Transformers (ViT) for the challenging exemplar-free scenario, with special focus on how to efficiently distill the knowledge of its crucial self-attention mechanism (SAM). Our work takes an initial step towards a surgical investigation of SAM for designing coherent continual learning methods in ViTs. We first carry out an evaluation of established continual learning regularization techniques. We then examine the effect of regularization when applied to two key enablers of SAM: (a) the contextualized embedding layers, for their ability to capture well-scaled representations with respect to the values, and (b) the prescaled attention maps, for carrying value-independent global contextual information. We depict the perks of each distilling strategy on two image recognition benchmarks (CIFAR100 and ImageNet-32) – while (a) leads to a better overall accuracy, (b) helps enhance the rigidity by maintaining competitive performances. Furthermore, we identify the limitation imposed by the symmetric nature of regularization losses. To alleviate this, we propose an asymmetric variant and apply it to the pooled output distillation (POD) loss adapted for ViTs. Our experiments confirm that introducing asymmetry to POD boosts its plasticity while retaining stability across (a) and (b). Moreover, we acknowledge low forgetting measures for all the compared methods, indicating that ViTs might be naturally inclined continual learners. 1
Keywords: Learning systems; Weight measurement; Image recognition; Surgery; Benchmark testing; Transformers; Stability analysis
|
Debora Gil, Jaume Garcia, Mariano Vazquez, Ruth Aris, & Guilleaume Houzeaux. (2008). Patient-Sensitive Anatomic and Functional 3D Model of the Left Ventricle Function. In 8th World Congress on Computational Mechanichs (WCCM8).
Abstract: Early diagnosis and accurate treatment of Left Ventricle (LV) dysfunction significantly increases the patient survival. Impairment of LV contractility due to cardiovascular diseases is reflected in its motion patterns. Recent advances in medical imaging, such as Magnetic Resonance (MR), have encouraged research on 3D simulation and modelling of the LV dynamics. Most of the existing 3D models [1] consider just the gross anatomy of the LV and restore a truncated ellipse which deforms along the cardiac cycle. The contraction mechanics of any muscle strongly depends on the spatial orientation of its muscular fibers since the motion that the muscle undergoes mainly takes place along the fibers. It follows that such simplified models do not allow evaluation of the heart electro-mechanical function and coupling, which has recently risen as the key point for understanding the LV functionality [2]. In order to thoroughly understand the LV mechanics it is necessary to consider the complete anatomy of the LV given by the orientation of the myocardial fibres in 3D space as described by Torrent Guasp [3].
We propose developing a 3D patient-sensitive model of the LV integrating, for the first time, the ven- tricular band anatomy (fibers orientation), the LV gross anatomy and its functionality. Such model will represent the LV function as a natural consequence of its own ventricular band anatomy. This might be decisive in restoring a proper LV contraction in patients undergoing pace marker treatment.
The LV function is defined as soon as the propagation of the contractile electromechanical pulse has been modelled. In our experiments we have used the wave equation for the propagation of the electric pulse. The electromechanical wave moves on the myocardial surface and should have a conductivity tensor oriented along the muscular fibers. Thus, whatever mathematical model for electric pulse propa- gation [4] we consider, the complete anatomy of the LV should be extracted.
The LV gross anatomy is obtained by processing multi slice MR images recorded for each patient. Information about the myocardial fibers distribution can only be extracted by Diffusion Tensor Imag- ing (DTI), which can not provide in vivo information for each patient. As a first approach, we have
Figure 1: Scheme for the Left Ventricle Patient-Sensitive Model.
computed an average model of fibers from several DTI studies of canine hearts. This rough anatomy is the input for our electro-mechanical propagation model simulating LV dynamics. The average fiber orientation is updated until the simulated LV motion agrees with the experimental evidence provided by the LV motion observed in tagged MR (TMR) sequences. Experimental LV motion is recovered by applying image processing, differential geometry and interpolation techniques to 2D TMR slices [5]. The pipeline in figure 1 outlines the interaction between simulations and experimental data leading to our patient-tailored model.
Keywords: Left Ventricle, Electromechanical Models, Image Processing, Magnetic Resonance.
|
Debora Gil, Jaume Garcia, Manuel Vazquez, Ruth Aris, & Guillaume Houzeaux. (2008). Patient-Sensitive Anatomic and Functional 3D Model of the Left Ventricle Function. In 8th World Congress on Computational Mechanichs (WCCM8)/5th European Congress on Computational Methods in Applied Sciences and Engineering (ECCOMAS 2008). Venezia (Italia).
Abstract: Early diagnosis and accurate treatment of Left Ventricle (LV) dysfunction significantly increases the patient survival. Impairment of LV contractility due to cardiovascular diseases is reflected in its motion patterns. Recent advances in medical imaging, such as Magnetic Resonance (MR), have encouraged research on 3D simulation and modelling of the LV dynamics. Most of the existing 3D models consider just the gross anatomy of the LV and restore a truncated ellipse which deforms along the cardiac cycle. The contraction mechanics of any muscle strongly depends on the spatial orientation of its muscular fibers since the motion that the muscle undergoes mainly takes place along the fibers. It follows that such simplified models do not allow evaluation of the heart electro-mechanical function and coupling, which has recently risen as the key point for understanding the LV functionality . In order to thoroughly understand the LV mechanics it is necessary to consider the complete anatomy of the LV given by the orientation of the myocardial fibres in 3D space as described by Torrent Guasp. We propose developing a 3D patient-sensitive model of the LV integrating, for the first time, the ven- tricular band anatomy (fibers orientation), the LV gross anatomy and its functionality. Such model will represent the LV function as a natural consequence of its own ventricular band anatomy. This might be decisive in restoring a proper LV contraction in patients undergoing pace marker treatment. The LV function is defined as soon as the propagation of the contractile electromechanical pulse has been modelled. In our experiments we have used the wave equation for the propagation of the electric pulse. The electromechanical wave moves on the myocardial surface and should have a conductivity tensor oriented along the muscular fibers. Thus, whatever mathematical model for electric pulse propa- gation [4] we consider, the complete anatomy of the LV should be extracted. The LV gross anatomy is obtained by processing multi slice MR images recorded for each patient. Information about the myocardial fibers distribution can only be extracted by Diffusion Tensor Imag- ing (DTI), which can not provide in vivo information for each patient. As a first approach, we have computed an average model of fibers from several DTI studies of canine hearts. This rough anatomy is the input for our electro-mechanical propagation model simulating LV dynamics. The average fiber orientation is updated until the simulated LV motion agrees with the experimental evidence provided by the LV motion observed in tagged MR (TMR) sequences. Experimental LV motion is recovered by applying image processing, differential geometry and interpolation techniques to 2D TMR slices [5]. The pipeline in figure 1 outlines the interaction between simulations and experimental data leading to our patient-tailored model.
Keywords: Left Ventricle; Electromechanical Models; Image Processing; Magnetic Resonance.
|
Lluis Garrido, M.Guerrieri, & Laura Igual. (2015). Image Segmentation with Cage Active Contours. TIP - IEEE Transactions on Image Processing, 24(12), 5557–5566.
Abstract: In this paper, we present a framework for image segmentation based on parametrized active contours. The evolving contour is parametrized according to a reduced set of control points that form a closed polygon and have a clear visual interpretation. The parametrization, called mean value coordinates, stems from the techniques used in computer graphics to animate virtual models. Our framework allows to easily formulate region-based energies to segment an image. In particular, we present three different local region-based energy terms: 1) the mean model; 2) the Gaussian model; 3) and the histogram model. We show the behavior of our method on synthetic and real images and compare the performance with state-of-the-art level set methods.
Keywords: Level sets; Mean value coordinates; Parametrized active contours; level sets; mean value coordinates
|
Enric Marti, Jordi Regincos, Jaime Lopez-Krahe, & Juan J.Villanueva. (1993). Hand line drawing interpretation as three-dimensional objects. Signal Processing – Intelligent systems for signal and image understanding, 32(1-2), 91–110.
Abstract: In this paper we present a technique to interpret hand line drawings as objects in a three-dimensional space. The object domain considered is based on planar surfaces with straight edges, concretely, on ansextension of Origami world to hidden lines. The line drawing represents the object under orthographic projection and it is sensed using a scanner. Our method is structured in two modules: feature extraction and feature interpretation. In the first one, image processing techniques are applied under certain tolerance margins to detect lines and junctions on the hand line drawing. Feature interpretation module is founded on line labelling techniques using a labelled junction dictionary. A labelling algorithm is here proposed. It uses relaxation techniques to reduce the number of incompatible labels with the junction dictionary so that the convergence of solutions can be accelerated. We formulate some labelling hypotheses tending to eliminate elements in two sets of labelled interpretations. That is, those which are compatible with the dictionary but do not correspond to three-dimensional objects and those which represent objects not very probable to be specified by means of a line drawing. New entities arise on the line drawing as a result of the extension of Origami world. These are defined to enunciate the assumptions of our method as well as to clarify the algorithms proposed. This technique is framed in a project aimed to implement a system to create 3D objects to improve man-machine interaction in CAD systems.
Keywords: Line drawing interpretation; line labelling; scene analysis; man-machine interaction; CAD input; line extraction
|
Josep Llados, Jaime Lopez-Krahe, & Enric Marti. (1997). A system to understand hand-drawn floor plans using subgraph isomorphism and Hough transform. In Machine Vision and Applications (Vol. 10, pp. 150–158).
Abstract: Presently, man-machine interface development is a widespread research activity. A system to understand hand drawn architectural drawings in a CAD environment is presented in this paper. To understand a document, we have to identify its building elements and their structural properties. An attributed graph structure is chosen as a symbolic representation of the input document and the patterns to recognize in it. An inexact subgraph isomorphism procedure using relaxation labeling techniques is performed. In this paper we focus on how to speed up the matching. There is a building element, the walls, characterized by a hatching pattern. Using a straight line Hough transform (SLHT)-based method, we recognize this pattern, characterized by parallel straight lines, and remove from the input graph the edges belonging to this pattern. The isomorphism is then applied to the remainder of the input graph. When all the building elements have been recognized, the document is redrawn, correcting the inaccurate strokes obtained from a hand-drawn input.
Keywords: Line drawings – Hough transform – Graph matching – CAD systems – Graphics recognition
|
Sounak Dey, Anguelos Nicolaou, Josep Llados, & Umapada Pal. (2016). Local Binary Pattern for Word Spotting in Handwritten Historical Document. In Joint IAPR International Workshops on Statistical Techniques in Pattern Recognition (SPR) and Structural and Syntactic Pattern Recognition (SSPR) (pp. 574–583). LNCS.
Abstract: Digital libraries store images which can be highly degraded and to index this kind of images we resort to word spotting as our information retrieval system. Information retrieval for handwritten document images is more challenging due to the difficulties in complex layout analysis, large variations of writing styles, and degradation or low quality of historical manuscripts. This paper presents a simple innovative learning-free method for word spotting from large scale historical documents combining Local Binary Pattern (LBP) and spatial sampling. This method offers three advantages: firstly, it operates in completely learning free paradigm which is very different from unsupervised learning methods, secondly, the computational time is significantly low because of the LBP features, which are very fast to compute, and thirdly, the method can be used in scenarios where annotations are not available. Finally, we compare the results of our proposed retrieval method with other methods in the literature and we obtain the best results in the learning free paradigm.
Keywords: Local binary patterns; Spatial sampling; Learning-free; Word spotting; Handwritten; Historical document analysis; Large-scale data
|
David Fernandez, Pau Riba, Alicia Fornes, & Josep Llados. (2014). On the Influence of Key Point Encoding for Handwritten Word Spotting. In 14th International Conference on Frontiers in Handwriting Recognition (pp. 476–481).
Abstract: In this paper we evaluate the influence of the selection of key points and the associated features in the performance of word spotting processes. In general, features can be extracted from a number of characteristic points like corners, contours, skeletons, maxima, minima, crossings, etc. A number of descriptors exist in the literature using different interest point detectors. But the intrinsic variability of handwriting vary strongly on the performance if the interest points are not stable enough. In this paper, we analyze the performance of different descriptors for local interest points. As benchmarking dataset we have used the Barcelona Marriage Database that contains handwritten records of marriages over five centuries.
Keywords: Local descriptors; Interest points; Handwritten documents; Word spotting; Historical document analysis
|
Marçal Rusiñol, J. Chazalon, & Jean-Marc Ogier. (2016). Filtrage de descripteurs locaux pour l'amélioration de la détection de documents. In Colloque International Francophone sur l'Écrit et le Document.
Abstract: In this paper we propose an effective method aimed at reducing the amount of local descriptors to be indexed in a document matching framework.In an off-line training stage, the matching between the model document and incoming images is computed retaining the local descriptors from the model that steadily produce good matches. We have evaluated this approach by using the ICDAR2015 SmartDOC dataset containing near 25000 images from documents to be captured by a mobile device. We have tested the performance of this filtering step by using ORB and SIFT local detectors and descriptors. The results show an important gain both in quality of the final matching as well as in time and space requirements.
Keywords: Local descriptors; mobile capture; document matching; keypoint selection
|
Pedro Martins, Paulo Carvalho, & Carlo Gatta. (2016). On the completeness of feature-driven maximally stable extremal regions. PRL - Pattern Recognition Letters, 74, 9–16.
Abstract: By definition, local image features provide a compact representation of the image in which most of the image information is preserved. This capability offered by local features has been overlooked, despite being relevant in many application scenarios. In this paper, we analyze and discuss the performance of feature-driven Maximally Stable Extremal Regions (MSER) in terms of the coverage of informative image parts (completeness). This type of features results from an MSER extraction on saliency maps in which features related to objects boundaries or even symmetry axes are highlighted. These maps are intended to be suitable domains for MSER detection, allowing this detector to provide a better coverage of informative image parts. Our experimental results, which were based on a large-scale evaluation, show that feature-driven MSER have relatively high completeness values and provide more complete sets than a traditional MSER detection even when sets of similar cardinality are considered.
Keywords: Local features; Completeness; Maximally Stable Extremal Regions
|
M. Altillawi, S. Li, S.M. Prakhya, Z. Liu, & Joan Serrat. (2024). Implicit Learning of Scene Geometry From Poses for Global Localization. ROBOTAUTOMLET - IEEE Robotics and Automation Letters, 9(2), 955–962.
Abstract: Global visual localization estimates the absolute pose of a camera using a single image, in a previously mapped area. Obtaining the pose from a single image enables many robotics and augmented/virtual reality applications. Inspired by latest advances in deep learning, many existing approaches directly learn and regress 6 DoF pose from an input image. However, these methods do not fully utilize the underlying scene geometry for pose regression. The challenge in monocular relocalization is the minimal availability of supervised training data, which is just the corresponding 6 DoF poses of the images. In this letter, we propose to utilize these minimal available labels (i.e., poses) to learn the underlying 3D geometry of the scene and use the geometry to estimate the 6 DoF camera pose. We present a learning method that uses these pose labels and rigid alignment to learn two 3D geometric representations ( X, Y, Z coordinates ) of the scene, one in camera coordinate frame and the other in global coordinate frame. Given a single image, it estimates these two 3D scene representations, which are then aligned to estimate a pose that matches the pose label. This formulation allows for the active inclusion of additional learning constraints to minimize 3D alignment errors between the two 3D scene representations, and 2D re-projection errors between the 3D global scene representation and 2D image pixels, resulting in improved localization accuracy. During inference, our model estimates the 3D scene geometry in camera and global frames and aligns them rigidly to obtain pose in real-time. We evaluate our work on three common visual localization datasets, conduct ablation studies, and show that our method exceeds state-of-the-art regression methods' pose accuracy on all datasets.
Keywords: Localization; Localization and mapping; Deep learning for visual perception; Visual learning
|
Mohammad Ali Bagheri, Qigang Gao, & Sergio Escalera. (2013). Logo recognition Based on the Dempster-Shafer Fusion of Multiple Classifiers. In 26th Canadian Conference on Artificial Intelligence (Vol. 7884, pp. 1–12). Springer Berlin Heidelberg.
Abstract: Best paper award
The performance of different feature extraction and shape description methods in trademark image recognition systems have been studied by several researchers. However, the potential improvement in classification through feature fusion by ensemble-based methods has remained unattended. In this work, we evaluate the performance of an ensemble of three classifiers, each trained on different feature sets. Three promising shape description techniques, including Zernike moments, generic Fourier descriptors, and shape signature are used to extract informative features from logo images, and each set of features is fed into an individual classifier. In order to reduce recognition error, a powerful combination strategy based on the Dempster-Shafer theory is utilized to fuse the three classifiers trained on different sources of information. This combination strategy can effectively make use of diversity of base learners generated with different set of features. The recognition results of the individual classifiers are compared with those obtained from fusing the classifiers’ output, showing significant performance improvements of the proposed methodology.
Keywords: Logo recognition; ensemble classification; Dempster-Shafer fusion; Zernike moments; generic Fourier descriptor; shape signature
|
A.Kesidis, & Dimosthenis Karatzas. (2014). Logo and Trademark Recognition. In D. Doermann, & K. Tombre (Eds.), Handbook of Document Image Processing and Recognition (Vol. D, pp. 591–646). Springer London.
Abstract: The importance of logos and trademarks in nowadays society is indisputable, variably seen under a positive light as a valuable service for consumers or a negative one as a catalyst of ever-increasing consumerism. This chapter discusses the technical approaches for enabling machines to work with logos, looking into the latest methodologies for logo detection, localization, representation, recognition, retrieval, and spotting in a variety of media. This analysis is presented in the context of three different applications covering the complete depth and breadth of state of the art techniques. These are trademark retrieval systems, logo recognition in document images, and logo detection and removal in images and videos. This chapter, due to the very nature of logos and trademarks, brings together various facets of document image analysis spanning graphical and textual content, while it links document image analysis to other computer vision domains, especially when it comes to the analysis of real-scene videos and images.
Keywords: Logo recognition; Logo removal; Logo spotting; Trademark registration; Trademark retrieval systems
|
Aura Hernandez-Sabate, David Rotger, & Debora Gil. (2008). Image-based ECG sampling of IVUS sequences. In Proc. IEEE Ultrasonics Symp. IUS 2008 (pp. 1330–1333).
Abstract: Longitudinal motion artifacts in IntraVascular UltraSound (IVUS) sequences hinders a properly 3D reconstruction and vessel measurements. Most of current techniques base on the ECG signal to obtain a gated pullback without the longitudinal artifact by using a specific hardware or the ECG signal itself. The potential of IVUS images processing for phase retrieval still remains little explored. In this paper, we present a fast forward image-based algorithm to approach ECG sampling. Inspired on the fact that maximum and minimum lumen areas are related to end-systole and end-diastole, our cardiac phase retrieval is based on the analysis of tissue density of mass along the sequence. The comparison between automatic and manual phase retrieval (0.07 ± 0.07 mm. of error) encourages a deep validation contrasting with ECG signals.
Keywords: Longitudinal Motion; Image-based ECG-gating; Fourier analysis
|
Onur Ferhat, & Fernando Vilariño. (2013). A Cheap Portable Eye-Tracker Solution for Common Setups. In 17th European Conference on Eye Movements.
Abstract: We analyze the feasibility of a cheap eye-tracker where the hardware consists of a single webcam and a Raspberry Pi device. Our aim is to discover the limits of such a system and to see whether it provides an acceptable performance. We base our work on the open source Opengazer (Zielinski, 2013) and we propose several improvements to create a robust, real-time system. After assessing the accuracy of our eye-tracker in elaborated experiments involving 18 subjects under 4 different system setups, we developed a simple game to see how it performs in practice and we also installed it on a Raspberry Pi to create a portable stand-alone eye-tracker which achieves 1.62° horizontal accuracy with 3 fps refresh rate for a building cost of 70 Euros.
Keywords: Low cost; eye-tracker; software; webcam; Raspberry Pi
|