130 results
Now showing1 - 10 of 130
- Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , Optimised preprocessing for automatic mouth gesture classificationMouth gestures are facial expressions in sign language, that do not refer to lip patterns of a spoken language. Research on this topic has been limited so far. The aim of this work is to automatically classify mouth gestures from video material by training a neural network. This could render time-consuming manual annotation unnecessary and help advance the field of automatic sign language translation. However, it is a challenging task due to the little data available as training material and the similarity of different mouth gesture classes. In this paper we focus on the preprocessing of the data, such as finding the area of the face important for mouth gesture recognition. Furthermore we analyse the duration of mouth gestures and determine the optimal length of video clips for classification. Our experiments show, that this can improve the classification results significantly and helps to reach a near human accuracy.Publicationtype: Conference Paper93 92 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , pinkIndexer – a universal indexer for pink-beam X-ray and electron diffraction snapshotsA crystallographic indexing algorithm,pinkIndexer, is presented for the analysisof snapshot diffraction patterns. It can be used in a variety of contexts includingmeasurements made with a monochromatic radiation source, a polychromaticsource or with radiation of very short wavelength. As such, the algorithm isparticularly suited to automated data processing for two emerging measurementtechniques for macromolecular structure determination: serial pink-beam X-raycrystallography and serial electron crystallography, which until now lackedreliable programs for analyzing many individual diffraction patterns fromcrystals of uncorrelated orientation. The algorithm requires approximateknowledge of the unit-cell parameters of the crystal, but not the wavelengthsassociated with each Bragg spot. The use ofpinkIndexeris demonstrated byobtaining 1005 lattices from a published pink-beam serial crystallography dataset that had previously yielded 140 indexed lattices. Additionally, in tests onexperimental serial crystallography diffraction data recorded with quasi-monochromatic X-rays and with electrons the algorithm indexed more patternsthan other programs tested.Publicationtype: Journal ArticleScopus© Citations 24 282 214 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Making metric learning algorithms invariant to transformations using a projection metric on Grassmann manifolds(2019-12-01); ; ; The requirement for suitable ways to measure the distance or similarity between data is omnipresent in machine learning, pattern recognition and data mining, but extracting such good metrics for particular problems is in general challenging. This has led to the emergence of metric learning ideas, which intend to automatically learn a distance function tuned to a specific task. In many tasks and data types, there are natural transformations to which the classification result should be invariant or insensitive. This demand and its implications are essential in many machine learning applications, and insensitivity to image transformations was in the first place achieved by using invariant feature vectors. In this paper, a new representation model on Grassmann manifolds for data points and a novel method for learning a Mahalanobis metric which uses the geodesic distance on Grassmann manifolds are proposed. In fact, we use an appropriate geodesic distance metric on the Grassmann manifolds, called projection metric, for measuring primary similarities between the new representations of the data points. This makes learning of the Mahalanobis metric invariant to similarity transforms and intensity changes, and therefore improve the performance. Experiments on face and handwritten digit datasets demonstrate that our proposed method yields performance improvements in a state-of-the-art metric learning algorithm.Publicationtype: Journal Article178 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , XGANDALF – extended gradient descent algorithm for lattice finding(Wiley-Blackwell, 2019-08-30); ; ; ; Serial crystallography records still diffraction patterns from single, randomly oriented crystals, then merges data from hundreds or thousands of them to form a complete data set. To process the data, the diffraction patterns must first be indexed, equivalent to determining the orientation of each crystal. A novel automatic indexing algorithm is presented, which in tests usually gives significantly higher indexing rates than alternative programs currently available for this task. The algorithm does not require prior knowledge of the lattice parameters but can make use of that information if provided, and also allows indexing of diffraction patterns generated by several crystals in the beam. Cases with a small number of Bragg spots per pattern appear to particularly benefit from the new approach. The algorithm has been implemented and optimized for fast execution, making it suitable for real-time feedback during serial crystallography experiments. It is implemented in an open-source C++ library and distributed under the LGPLv3 licence. An interface to it has been added to the CrystFEL software suite.Publicationtype: Journal ArticleScopus© Citations 71 500 227 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Classification of Mouth Gestures in German Sign Language using 3D Convolutional Neural NetworksAutomatic recognition of sign language gestures is becoming necessary with an increased interest into human-computer interaction in sign language as well as automatic translation from sign language. Most of the research on sign language recognition focuses on hand gesture recognition. However, there are also non-manual signals in sign language. Mouth gestures represent mouth shapes that add information to the hand gestures not related to spoken language visemes. For German Sign Language, mouth gesture recognition would be an important addition to manual gesture recognition. This research work evaluates the method 3D convolutional neural networks for recognising mouth gestures in German Sign Language. For the recognition of certain mouth gestures, temporal information is mandatory and the extraction of both spatial and temporal features by 3D convolutional networks makes the classification of all gestures easier. Our research work compares how different initialisations affect learning and classification by the network. We achieve an accuracy of around 68% on testing 10 classes of mouth gestures in German Sign Language. © 2019 IET Conference Publications. All rights reserved.Publicationtype: Conference Paper336 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Variance and symmetrical-based approach for optimal alignment of 3D model(2018-10); ; The concept of building 3D models, known as 3D reconstruction, already exists since the last few decades. However, by manually aligning the objects during acquisition phase does not guarantee that the output, the 3D models, will be perfectly aligned with the computer's world coordinate system. It mainly happens because in real world it is quite challenging to get perfect measurements, especially for the irregular objects. In this paper we address this problem by proposing a method to be used on the post processing phase of the 3D reconstruction process. The method is based on the variance and symmetricity of the object's point cloud which is acquired during acquisition. For the evaluation, we applied and evaluated the proposed method to both synthetic and reconstructed 3D models. The results are significant and show that the method capable of aligning the models to a fine resolution of 1' (one minute) angle.Publicationtype: Conference Paper132 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , An Automated Approach for Inner Segment/Outer Segment Defect Detection in Retinal SD-OCT Images(2018-08-01); ; ; There has been shown a strong association between the integrity of the inner/outer segment (IS/OS) junction and the visual acuity in various retinal diseases. We propose an automated method for IS/OS defect detection in optical coherence tomography images with a focus on resilience, particularly with epiretinal membranes present.50 eyes with epiretinal membranes were included in this study, and received retinal scans were acquired using spectral domain optical coherence tomography. The algorithm is based on the pixel value maxima above the retinal pigment epithelium. Summing up the pixel values row-wise to the lower part of the image leads to two maxima with respect to the IS and OS layer. The classification algorithm itself is divided into two parts, the polynomial interpolation and the Gabor filtering. Performing both parts yield to the classification IS/OS defect. In order to quantify the algorithmic performance of the proposed method, the resulting classification was compared against corresponding gold standard data. Algorithmically classified pixels were compared to this gold standard. The algorithm reached a mean performance of 96.77% sensitivity and 99.09% specificity. We demonstrate that the automatic classification of IS/OS defects in optical coherence tomography is possible with high sensitivity and specificity.Publicationtype: Journal Article129 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Surface reconstruction from image space adjacency of lines using breadth-first plane search(2016-06-08); In this paper, we propose a novel multi-view method for surface reconstruction from matched line segments with applications to robotic mapping and image-based rendering. Starting from 3D line segments and poses obtained via Line-SLAM, we project segments from multiple frames into keyframes for image-space analysis. For each keyframe, a grid of image faces is created by optimized intersection of the segment projection lines. These faces define a segment adjacency graph, wherein we perform our Breadth-First Plane Search (BFPS). The found plane hypotheses are merged maximally with respect to a structure-preserving criterion by growing coplanar regions across the graph. Hypotheses violating the visibility constraint are discarded based on fast per-face and mostly nongeometrical evaluation of the scene and image graph. Finally, each image face gets back-projected onto an optimal plane to obtain a 3D surface model. The presented system is a complete and automatic solution suitable for mapping an environment in real-time scenarios like robotic exploration. We demonstrate the performance of our algorithm on several indoor scenes of varying complexity. Compared to a pure 3D analysis of segments, we see a speed-up by one to almost two orders of magnitude, while still improving on reconstruction accuracy.Publicationtype: Conference Paper78 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Automatisierte Klassifikation retinaler Gefäße in konfokalen Laser-Scanning-Ophthalmoskopie Aufnahmen(2015-10); ; ; Publicationtype: Conference Poster not in Proceedings48 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Sparse reconstruction of liver cirrhosis from monocular mini-laparoscopic sequencesMini-laparoscopy is a technique which is used by clinicians to inspect the liver surface with ultra-thin laparoscopes. However, so far no quantitative measures based on mini-laparoscopic sequences are possible. This paper presents a Structure from Motion (SfM) based methodology to do 3D reconstruction of liver cirrhosis from mini-laparoscopic videos. The approach combines state-of-the-art tracking, pose estimation, outlier rejection and global optimization to obtain a sparse reconstruction of the cirrhotic liver surface. Specular reflection segmentation is included into the reconstruction framework to increase the robustness of the reconstruction. The presented approach is evaluated on 15 endoscopic sequences using three cirrhotic liver phantoms. The median reconstruction accuracy ranges from 0.3 mm to 1 mm.Publicationtype: Conference Paper58