| 1. |
- Hast, Anders, 1966-, et al.
(författare)
-
Ensembles and Cascading of Embedded Prototype Subspace Classifiers
- 2020
-
Ingår i: Journal of WSCG. - 1213-6972 .- 1213-6964. ; 28:1/2, s. 89-95
-
Tidskriftsartikel (refereegranskat)abstract
- Deep learning approaches suffer from the so called interpretability problem and can therefore be very hard to visualise. Embedded Prototype Subspace Classifiers is one attempt in the field of explainable AI, which is both fast and efficient since it does not require repeated learning epochs and has no hidden layers. In this paper we investigate how ensembles and cascades of ensembles perform on some popular datasets. The focus is on handwritten data such as digits, letters and signs. It is shown how cascading can be efficiently implemented in order to both increase accuracy as well as speed up the classification.
|
|
| 2. |
- Hast, Anders, 1966-
(författare)
-
Magnitude of Semicircle Tiles in Fourier-space : A Handcrafted Feature Descriptor for Word Recognition using Embedded Prototype Subspace Classifiers
- 2022
-
Ingår i: Journal of WSCG. - 1213-6972 .- 1213-6964. ; 30:1-2, s. 82-90
-
Tidskriftsartikel (refereegranskat)abstract
- The purpose of this paper is to in detail describe and analyse a Fourier based handcrafted descriptor for word recognition. Especially, it is discussed how the Variability in the results can be analysed and visualised. This efficiency of the descriptor is evaluated for the use with embedded prototype subspace classifiers for handwritten word recognition. Nonetheless, it can be used with any classifier for any purpose. An hierarchical composition of discrete semicircles in the Fourier-space is proposed and it will will be show how this compares to Gabor filters, which can be used to extract edges in an image. In comparison to Histogram of Oriented Gradients, the proposed feature descriptor performs better in this scenario. Compression using PCA turns out to be able to increase both the F1-score as well as decreasing the Variability.
|
|
| 3. |
|
|
| 4. |
|
|
| 5. |
- Hast, Anders, 1966-, et al.
(författare)
-
Word Recognition using Embedded Prototype Subspace Classifiers on a new Imbalanced Dataset
- 2021
-
Ingår i: Journal of WSCG. - : University of West Bohemia. - 1213-6972 .- 1213-6964. ; 29:1-2, s. 39-47
-
Tidskriftsartikel (refereegranskat)abstract
- This paper presents an approach towards word recognition based on embedded prototype subspace classification. The purpose of this paper is three-fold. Firstly, a new dataset for word recognition is presented, which is extracted from the Esposalles database consisting of the Barcelona cathedral marriage records. Secondly, different clustering techniques are evaluated for Embedded Prototype Subspace Classifiers. The dataset, containing 30 different classes of words is heavily imbalanced, and some word classes are very similar, which renders the classification task rather challenging. For ease of use, no stratified sampling is done in advance, and the impact of different data splits is evaluated for different clustering techniques. It will be demonstrated that the original clustering technique based on scaling the bandwidth has to be adjusted for this new dataset. Thirdly, an algorithm is therefore proposed that finds k clusters, striving to obtain a certain amount of feature points in each cluster, rather than finding some clusters based on scaling the Silverman’s rule of thumb. Furthermore, Self Organising Maps are also evaluated as both a clustering and embedding technique.
|
|
| 6. |
- Napieralla, Jonah, et al.
(författare)
-
Ultrawide field of view by curvilinear projection methods
- 2020
-
Ingår i: Journal of WSCG. - : Vaclav Skala Union Agency. - 1213-6972 .- 1213-6964. ; 28:1-2, s. 163-167
-
Tidskriftsartikel (refereegranskat)abstract
- The rectilinear Perspective projection produces natural-looking results on the condition that the degree of field of view (FoV) is narrow, as raising it causes an exponential increase in visual distortion. Curvilinear perspective projection methods that counter this issue exist in photography, but unlike the rectilinear Perspective projection, these are neither used nor technically documented in computer graphics. This paper contributes by presenting results from a perceptual experiment comparing the rectilinear Perspective projection method to the curvilinear Panini and Stereographic projection methods. These are commonly used with wide-angle lenses in photography and have been digitally recreated for use in computer graphics. The experiment shows a clear preference for the two curvilinear projection methods at high degrees of FoV, as not a single participant prefers the rectilinear Perspective projection at degrees of FoV approaching the human breadth of vision. © 2020, Vaclav Skala Union Agency. All rights reserved.
|
|
| 7. |
- Nysjö, Fredrik, 1985-
(författare)
-
Clustered Grid Cell Data Structure for Isosurface Rendering
- 2020
-
Ingår i: Journal of WSCG. - 1213-6972 .- 1213-6964. ; 28:1-2, s. 9-17
-
Tidskriftsartikel (refereegranskat)abstract
- Active grid cells in scalar volume data are typically identified by many isosurface rendering methods when extracting another representation of the data for rendering. However, the use of grid cells themselves as rendering primitives is not extensively explored in the literature. In this paper, we propose a cluster-based data structure for storing the data of active grid cells for fast cell rasterisation via billboard splatting. Compared to previous cell rasterisation approaches, eight corner scalar values are stored with each active grid cell, so that the full volume data is not required during rendering. The grid cells can be quickly extracted and use about 37 percent memory compared to a typical efficient mesh-based representation, while supporting large grid sizes. We present further improvements such as a visibility buffer for cluster culling and EWA-based interpolation of attributes such as normals. We also show that our data structure can be used for hybrid ray tracing or path tracing to compute global illumination.
|
|
| 8. |
|
|
| 9. |
- Nysjö, Johan, et al.
(författare)
-
BoneSplit - A 3D Texture Painting Tool for Interactive Bone Separation in CT Images
- 2015
-
Ingår i: Journal of WSCG. - 1213-6972 .- 1213-6964. ; 23:2, s. 157-166
-
Tidskriftsartikel (refereegranskat)abstract
- We present an efficient interactive tool for separating collectively segmented bones and bone fragments in 3D computed tomography (CT) images. The tool, which is primarily intended for virtual cranio-maxillofacial (CMF) surgery planning, combines direct volume rendering with an interactive 3D texture painting interface to enable quick identification and marking of individual bone structures. The user can paint markers (seeds) directly on the rendered bone surfaces as well as on individual CT slices. Separation of the marked bones is then achieved through the random walks segmentation algorithm, which is applied on a graph constructed from the collective bone segmentation. The segmentation runs on the GPU and can achieve close to real-time update rates for volumes as large as 512^3. Segmentation editing can be performed both in the random walks segmentation stage and in a separate post-processing stage using a local 3D editing tool. In a preliminary evaluation of the tool, we demonstrate that segmentation results comparable with manual segmentations can be obtained within a few minutes.
|
|
| 10. |
- Zhang, Yuchong, 1994, et al.
(författare)
-
Automated microwave tomography (Mwt) image segmentation: State-of-the-art implementation and evaluation
- 2020
-
Ingår i: Journal of WSCG. - 1213-6972 .- 1213-6964. ; 2020:2020, s. 126-136
-
Tidskriftsartikel (refereegranskat)abstract
- Inspired by the high performance in image-based medical analysis, this paper explores the use of advanced segmentation techniques for industrial Microwave Tomography (MWT). Our context is the visual analysis of moisture levels in porous foams undergoing microwave drying. We propose an automatic segmentation technique—MWT Segmentation based on K-means (MWTS-KM) and demonstrate its efficiency and accuracy for industrial use. MWTS-KM consists of three stages: image augmentation, grayscale conversion, and K-means implementation. To estimate the performance of this technique, we empirically benchmark its efficiency and accuracy against two well-established alternatives: Otsu and K-means. To elicit performance data, three metrics (Jaccard index, Dice coefficient and false positive) are used. Our results indicate that MWTS-KM outperforms the well-established Otsu and K-means, both in visually observable and objectively quantitative evaluation.
|
|
