SciTePress - Affiliation Profile

Research.Publish.Connect.

*Please fill out at least one Field. *Value must be an number!

Title:
ISBN:
Year:
Acronym:
Subject:

Advanced Search Proceedings Search

If you're looking for an exact phrase use quotation marks on text fields.

*Please fill out at least one Field.

Title:
Author:
Affiliation:
Subject:

Advanced Search Papers Search

If you're looking for an exact phrase use quotation marks on text fields.

*Please fill out at least one Field.

Name:
Affiliation:
Country:
Conference:
Subject:

Advanced Search Authors Search

If you're looking for an exact phrase use quotation marks on text fields.

*Please fill out at least one Field.

Name:
Country:
Subject:

Advanced Search Affiliations Search

If you're looking for an exact phrase use quotation marks on text fields.

Proceedings

Proceedings Search *Please fill out at least one Field. *Value must be an number!

Title:
ISBN:
Year:
Acronym:
Subject:

Advanced Search Proceedings Search

If you're looking for an exact phrase use quotation marks on text fields.

Papers

Papers Search *Please fill out at least one Field.

Title:
Author:
Affiliation:
Subject:

Advanced Search Papers Search

If you're looking for an exact phrase use quotation marks on text fields.

Authors

Authors Search *Please fill out at least one Field.

Name:
Affiliation:
Country:
Conference:
Subject:

Advanced Search Authors Search

If you're looking for an exact phrase use quotation marks on text fields.

Advanced Search

Affiliation

Select a Subject
Refine Ontology Subject

University of Applied Sciences Hagenberg, Austria

Context (please use complete words only)

Field

Sort by

16 Papers

Generative Adversarial Network Synthesis for Improved Deep Learning Model Training of Alpine Plants with Fuzzy Structures

Topics: Segmentation and Grouping; Features Extraction; Object Detection and Localization; Assistive Computer Vision

Authors: Christoph Praschl ¹ , Roland Kaiser ² and Gerald Zwettler ^{1

;

3}

Affiliations: ¹ Research Group Advanced Information Systems and Technology, Research and Development Department, University of Applied Sciences Upper Austria, Softwarepark 11, Hagenberg, 4232, Austria , ² ENNACON Environment Nature Consulting KG, Altheim 13, Feldkirchen bei Mattighofen, Austria , ³ Department of Software Engineering, School of Informatics, Communications and Media, University of Applied Sciences Upper Austria, Softwarepark 11, Hagenberg, 4232, Austria

Digital Object Identifier: 10.5220/0011607100003417

Proceedings of VISAPP 2023/VISIGRAPP, ISBN: 978-989-758-634-7

Quick Abstract

Deep learning approaches are highly influenced by two factors, namely the complexity of the task and the size of the training data set. In terms of both, the extraction of features of low-stature alpine plants represents a challenging domain due to their fuzzy appearance, a great structural variety in plant organs and the high effort associated with acquiring high-quality training data for such plants. For this reason, this study proposes an approach for training deep learning models in the context of alpine vegetation based on a combination of real-world and artificial data synthesised using Generative Adversarial Networks. The evaluation of this approach indicates that synthetic data can be used to increase the size of training data sets. With this at hand, the results and robustness of deep learning models are demonstrated with a U-Net segmentation model. The evaluation is carried out using a cross-validation for three alpine plants, namely Soldanella pusilla, Gnaphalium supinum, and Euphrasia minima. Improved segmentation accuracy was achieved for the latter two species. Dice Scores of 24.16% vs 26.18% were quantified for Gnaphalium with 100 real-world training images. In the case of Euphrasia, Dice Scores improved from 33.56% to 42.96% using only 20 real-world training images.

Three-step Approach for Localization, Instance Segmentation and Multi-facet Classification of Individual Logs in Wooden Piles

Topics: Ensemble Methods; Industry Related Applications; Deep Learning and Neural Networks; Image and Video Analysis and Understanding

Authors: Christoph Praschl ¹ and Gerald Adam Zwettler ^{2

;

1}

Affiliations: ¹ Research Group Advanced Information Systems and Technology (AIST), University of Applied Sciences Upper Austria, Softwarepark 11, 4232 Hagenberg, Austria , ² Department of Software Engineering, School of Informatics, Communications and Media, University of Applied Sciences Upper Austria, Softwarepark 11, 4232 Hagenberg, Austria

Digital Object Identifier: 10.5220/0010892100003122

Proceedings of ICPRAM 2022, ISBN: 978-989-758-549-4

Quick Abstract

The inspection of products and the assessment of quality is connected with high costs and time effort in many industrial domains. This also applies to the forestry industry. Utilizing state-of-the-art deep learning models allows the analysis automation of wooden piles in a vision-based manner. In this work a three-step approach is presented for the localization, segmentation and multi-facet classification of individual logs based on a client/server architecture allowing to determine the quality, volume and like this the value of a wooden pile based on a smartphone application. Using multiple YOLOv4 and U-NET models leads to a client-side log localization accuracy of 82.9% with low storage requirements of 23 MB and a server-side log detection accuracy of 94.1%, together with a log type classification accuracy of 95% and 96% according to the quality assessment of spruce logs. In addition, the trained segmentation model reaches an accuracy of 89%.

Generic User-guided Interaction Paradigm for Precise Post-slice-wise Processing of Tomographic Deep Learning Segmentations Utilizing Graph Cut and Graph Segmentation

Topics: Segmentation and Grouping; Medical Image Applications; Human and Computer Interaction; 3D Deep Learning

Authors: Gerald A. Zwettler ^{1

;

2

;

3} , Werner Backfrieder ² , Ronald A. Karwoski ¹ and David R. Holmes Iii ¹

Affiliations: ¹ Biomedical Analytics and Computational Engineering Lab, Department of Physiology and Biomedical Engineering, Mayo Clinic College of Medicine, 200 First St. SW, 55905 Rochester, MN, U.S.A. , ² Medical Informatics, Department of Software Engineering, University of Applied Sciences Upper Austria, Softwarepark 11, 4232 Hagenberg, Austria , ³ Research Group Advanced Information Systems and Technology (AIST), Department of Software Engineering, University of Applied Sciences Upper Austria, Softwarepark 11, 4232 Hagenberg, Austria

Digital Object Identifier: 10.5220/0010190702350244

Proceedings of VISAPP 2021/VISIGRAPP, ISBN: 978-989-758-488-6

Quick Abstract

State of the art deep learning (DL) manifested in image processing as an accurate segmentation method. Nevertheless, its black-box nature hardly allows user interference. In this paper, we present a generic Graph cut (GC) and Graph segmentation (GS) approach for user-guided interactive post-processing of segmentations resulting from DL. The GC fitness function incorporates both, the original image characteristics and DL segmentation results, combining them with weights optimized by evolution strategy optimization. To allow for accurate user-guided processing, the fore- and background seeds of the Graph cut are automatically selected from the DL segmentations, but implementing effective features for expert input for adaptions of position and topology. The seamless integration of DL with GC/GS leads to marginal trade-off in quality, namely Jaccard (JI) 1.3% for automated GC and JI 0.46% for GS only. Yet, in specific areas where a well-trained DL model may potentially fail, precise adaptions at a low demand for user-interaction become feasible and thus even outperforming the original DL results. The potential of GC/GS is shown running on ground- truth seeds thereby outperforming DL by 0.44% JI for the GC and even by 1.16% JI for the GS. Iterative slice- by-slice progression of the post-processed and improved results keeps the demand for user-interaction low.

Three-step Alignment Approach for Fitting a Normalized Mask of a Person Rotating in A-Pose or T-Pose Essential for 3D Reconstruction based on 2D Images and CGI Derived Reference Target Pose

Topics: Image-Based Modeling and 3D Reconstruction; Image Registration; Stereo Vision and Structure from Motion; Computational Photography

Authors: Gerald Adam Zwettler ^{1

;

2} , Christoph Praschl ² , David Baumgartner ² , Tobias Zucali ³ , Dora Turk ³ , Martin Hanreich ² and Andreas Schuler ^{4

;

2}

Affiliations: ¹ Department of Software Engineering, School of Informatics, Communications and Media, University of Applied Sciences Upper Austria, Softwarepark 11, 4232 Hagenberg, Austria , ² Research Group Advanced Information Systems and Technology (AIST), University of Applied Sciences Upper Austria, Softwarepark 11, 4232 Hagenberg, Austria , ³ AMB GmbH, amb-technology.ai, Hafenstraße 47-51 4020 Linz, Austria , ⁴ Department of Bio and Medical Informatics, School of Informatics, Communications and Media, University of Applied Sciences Upper Austria, Softwarepark 11, 4232 Hagenberg, Austria

Digital Object Identifier: 10.5220/0010194102810292

Proceedings of VISAPP 2021/VISIGRAPP, ISBN: 978-989-758-488-6

Quick Abstract

The 3D silhouette reconstruction of a human body rotating in front of a monocular camera system is a very challenging task due to elastic deformation and positional mismatch from body motion. Nevertheless, knowledge of the 3D body shape is a key information for precise determination of one’s clothing sizes, e.g. for precise shopping to reduce the number of return shipments in online retail. In this paper a novel three step alignment process is presented, utilizing As-Rigid-As-Possible (ARAP) transformations to normalize the body joint skeleton derived from OpenPose with a CGI rendered reference model in A- or T-pose. With further distance-map accelerated registration steps, positional mismatches and inaccuracies from the OpenPose joint estimation are compensated thus allowing for 3D silhouette reconstruction of a moving and elastic object without the need for sophisticated statistical shape models. Tests on both, artificial and real-world data, generally proof the practicability of this approach with all three alignment/registration steps essential and adequate for 3D silhouette reconstruction data normalization.

Pre- and Post-processing Strategies for Generic Slice-wise Segmentation of Tomographic 3D Datasets Utilizing U-Net Deep Learning Models Trained for Specific Diagnostic Domains

Topics: Segmentation and Grouping; Machine Learning Technologies for Vision; Medical Image Applications; Human and Computer Interaction; Deep Learning for Visual Understanding

Authors: Gerald A. Zwettler ^{1

;

2} , Werner Backfrieder ³ and David R. Holmes III ¹

Affiliations: ¹ Biomedical Analytics and Computational Engineering Lab, Department of Physiology and Biomedical Engineering, Mayo Clinic College of Medicine, 200 First St. SW, 55905 Rochester, MN, U.S.A. , ² Research Group Advanced Information Systems and Technology (AIST), Department of Software Engineering, University of Applied Sciences Upper Austria, Softwarepark 11, 4232 Hagenberg, Austria , ³ Medical Informatics, Department of Software Engineering, University of Applied Sciences Upper Austria, Softwarepark 11, 4232 Hagenberg, Austria

Digital Object Identifier: 10.5220/0008932100660078

Proceedings of VISAPP 2020/VISIGRAPP, ISBN: 978-989-758-402-2

Quick Abstract

An automated and generally applicable method for segmentation is still in focus of medical image processing research. Since a few years artificial inteligence methods show promising results, especially with widely available scalable Deep Learning libraries. In this work, a five layer hybrid U-net is developed for slice-by-slice segmentation of liver data sets. Training data is taken from the Medical Segmentation Decathlon database, providing 131 fully segmented volumes. A slice-oriented segmentation model is implemented utilizing deep learning algorithms with adaptions for variable parenchyma shape along the stacking direction and similarities between adjacent slices. Both are transformed for coronal and sagittal views. The implementation is on a GPU rack with TensorFlow and Keras. For a quantitative measure of segmentation accuracy, standardized volume and surface metrics are used. Results DSC=97.59, JI=95.29 and NSD=99.37 show proper segmentation comparable to 3D U-Nets and other state of the art. The development of a 2D-slice oriented segmentation is justified by short training time and less complexity and therefore massively reduced memory consumption. This work manifests the high potential of AI methods for general use in medical segmentation as fully- or semi-automated tool supervised by the expert user.

Who Wants to Use an Augmented Reality Shopping Assistant Application?

Topics: Recommendation Systems ; Shopping Assistants ; Augmented Reality

Authors: Daniel Mora ¹ , Robert Zimmermann ² , Douglas Cirqueira ³ , Marija Bezbradica ³ , Markus Helfert ⁴ , Andreas Auinger ² and Dirk Werth ¹

Affiliations: ¹ Artificial Intelligence Lab, AWS Institute for Digitized Products and Processes, Saarbrücken, Germany , ² Digital Business Management, University of Applied Sciences Upper Austria, Styer, Austria , ³ School of Computing, Dublin City University, Dublin, Ireland , ⁴ Innovation Value Institute, Maynooth University, Maynooth, Ireland

Digital Object Identifier: 10.5220/0010214503090318

Proceedings of WUDESHI-DR 2020, ISBN: 978-989-758-480-0

Quick Abstract

Brick-and-mortar retailers need to stay competitive to the convenience provided by online channels. Technologies, such as personalized shopping assistants on smartphones can empower customers in-store towards a similar experience as in an online scenario. For instance, an augmented reality shopping assistance application with explainable recommendations (XARSAA) can mimic the behavior of recommender systems in personalizing offers to consumers in physical shops. However, before deploying such technologies, it is essential that retailers get to know the demographics of their customer base. Existing literature rarely addresses the influence of customers demographics towards XARSAA technologies. Therefore, we follow a design science approach, and develop an instantiation of a XARSAA artifact, which is artificially evaluated through a controlled online user experiment with 315 participants. Results illustrate multiple demographics which influence customers attitude towards an augmented reality shopping assistant application in brick-and-mortar stores. Additionally, we provide insights into the design of such technology to guide researchers in its implementation.

Contextual Design in Industrial Settings: Experiences and Recommendations

Topics: HCI on Enterprise Information Systems; Human Factors; Requirements Analysis And Management; Guidelines, Principles, Patterns and Standards

Authors: Mirjam Augstein , Thomas Neumayr , Sebastian Pimminger , Christine Ebner , Josef Altmann and Werner Kurschl

Affiliation: University of Applied Sciences Upper Austria, Austria

Digital Object Identifier: 10.5220/0006674904290440

Proceedings of ICEIS 2018, ISBN: 978-989-758-298-1

Quick Abstract

The Contextual Design (CD) methodology offers a framework for planning and implementing a user-centered design process throughout all project phases. It is team-based and was designed especially for interdisciplinary teams. The application of CD is particularly profitable in projects confronted with implicit requirements and hidden factors of influence. In contrast to many other design and evaluation methods such as focus groups or usability tests, CD does not take users out of their everyday setting and more easily reveals important design issues and contextual influences like users’ motivation, values, emotions or real-time interruptions. Despite these advantages, CD is often not used due to high costs in terms of time and effort. This paper provides a report on experiences with CD in two research projects in the industry domain. It is intended to help other researchers to plan and implement a CD process in industrial settings and benefit from our lessons learned.

Related Ontology Subjects/Areas/Topics (8)

Enterprise Information Systems	Human Factors	Physiological Computing Systems
Guidelines, Principles, Patterns and Standards	Human-Computer Interaction	Requirements Analysis And Management
HCI on Enterprise Information Systems	Information Systems Analysis and Specification

Tangible Interaction for Simple 3D Interaction Tasks: Comparing Device-In-Hand and Hand-In-Device Scenarios

Topics: User Behaviour Analysis; Haptic and Tangible Devices

Authors: Mirjam Augstein ¹ , Thomas Neumayr ² , Stephan Vrecer ² , Werner Kurschl ³ and Josef Altmann ¹

Affiliations: ¹ Communication and Knowledge Media, University of Applied Sciences Upper Austria, Hagenberg and Austria , ² Research & Development, University of Applied Sciences Upper Austria, Hagenberg and Austria , ³ Human-Centered Computing, University of Applied Sciences Upper Austria, Hagenberg and Austria

Digital Object Identifier: 10.5220/0006883300230033

Proceedings of CHIRA 2018, ISBN: 978-989-758-328-5

Quick Abstract

Although recently, touch-based input with a reduced amount of haptic guidance gained popularity, traditional tangible input devices like mice or joysticks are still indispensable of everyday human-computer interaction. Most traditional tangible devices are implemented in so-called device-in-hand settings. There, the user’s hand grabs the device and the device-hand system is then moved to trigger an input activity. Thus, the user’s hand posture usually stays relatively stable. Hand-in-device approaches are an alternative form of tangible interaction settings where the user’s hand moves within a tangible device. Such a setting differs considerably as i) the user’s hand posture is more flexible, and ii) the device itself is stable while the interacting hand moves. This paper describes a comparative study on users’ interaction performance with device-in-hand and hand-in-device settings for simple 3D interaction tasks. Further, it contributes to the body of knowledge on favourable interaction directions (left or right).

The Role of Haptics in User Input for Simple 3D Interaction Tasks - An Analysis of Interaction Performance and User Experience

Topics: Fundamental of Haptic Interactions; Haptic Feedback and Control; Input and Output Modalities; User Performance Evaluation; Usability and User Experience

Authors: Mirjam Augstein , Thomas Neumayr , Stephan Vrecer , Werner Kurschl and Josef Altmann

Affiliation: University of Applied Sciences Upper Austria, Austria

Digital Object Identifier: 10.5220/0006580500260037

Proceedings of HUCAPP 2018/VISIGRAPP, ISBN: 978-989-758-288-2

Quick Abstract

Traditionally, input devices allowed for at least a certain degree of haptic experience by involving direct physical contact between user and device. Recently, touchless interaction gained popularity through readily available, cheap devices like the Leap motion controller or Microsoft Kinect. Usually, these devices support more than two degrees of freedom and are thus especially suitable for interaction tasks in a three-dimensional space. However, besides the high potential that lies within touchless input techniques, they also involve new challenges (e.g., lack of borders and natural haptic guidance). In this paper, we aim at the identification of potentials and limitations inherent to three different input techniques that involve a varying amount of haptics (i.e., touchful, touchless and semi-touchless input). We present a study conducted with 25 users that focuses on simple input tasks in a 3D interaction space and analyzes objective interaction performance metrics (e.g., regularity or time) and subjective User Experience aspects (e.g., dependability or efficiency). It reveals parallels as well as contrasts between the users’ actual interaction performance and perceived UX (e.g., several metrics suggested haptic input to outperform touchless input while differences regarding UX were not significant). The results are intended to inform other researchers when designing interactive environments.

Report Optimization using Visual Search Strategies - An Experimental Study with Eye Tracking Technology

Topics: Data Management and Knowledge Representation; Graph Visualization; Interpretation and Evaluation Methods; Perception and Cognition in Visualization; Usability Studies and Visualization

Authors: Lisa Falschlunger , Christoph Eisl , Heimo Losbichler and Elisabeth Grabmann

Affiliation: University of Applied Sciences Upper Austria, Austria

Digital Object Identifier: 10.5220/0005251702090218

Proceedings of IVAPP 2015/VISIGRAPP, ISBN: 978-989-758-088-8

Quick Abstract

The success of visualisations is determined by the ability of users to retrieve relevant information in an effective and efficient way. The way in which information is perceived can be analysed by examining visual search strategies of users. Visual search strategies in graphical representations however, are individual and have not been well explored up to now. Recent studies show that eye tracking experiments help in gaining new insights into these strategies. Apart from error rates and task completion times, eye tracking focuses on the way observers of visualisations read and make sense of the presented stimulus. In this way sequential strategies can be analysed, compared and used in order to optimize graphical layouts. In this study we use the approach of Parallel Scan Path visualisation in combination with Levenshtein Distance to determine similarities between search strings when viewing graphical representations in standardized business communication. This study shows a positive correlation between search strategies and task completion time and allows the evaluation of different design layouts. Positive significant effects can be detected when examining experience (with respect to standardized and repetitive reporting) and layout optimization (with respect to graphical representations and page layout). Optimal search strategies can be identified when users are experienced and using an optimized layout.

Related Ontology Subjects/Areas/Topics (8)

Abstract Data Visualization	General Data Visualization	Perception and Cognition in Visualization
Computer Vision, Visualization and Computer Graphics	Graph Visualization	Usability Studies and Visualization
Data Management and Knowledge Representation	Interpretation and Evaluation Methods