Refine
Year of publication
- 2023 (38) (remove)
Document Type
- Article (unreviewed) (38) (remove)
Is part of the Bibliography
- yes (38)
Keywords
- Deep Learning (3)
- Export (3)
- Ransomware (3)
- Social Engineering (3)
- Advanced Footwear Technology (2)
- Biomechanics (2)
- Digitalisierung (2)
- Exercise Science (2)
- ISO (2)
- IT Governance (2)
- Public Service Models (2)
- Wärmepumpe (2)
- deep learning (2)
- Analytics (1)
- Automation (1)
- Bauökologie (1)
- Bildung (1)
- Bildungsmanagement (1)
- Bogenmodell (1)
- COVID-19 (1)
- Cloud (1)
- Computergestützte Fertigung (1)
- Dashcam (1)
- Datenschutz (1)
- Deep Leaning (1)
- Deep Reinforcement Learning (1)
- DenseNet (1)
- DenseNet201 (1)
- Digitalisierte Arbeitswelt (1)
- Disease Detection (1)
- Disease detection (1)
- ECA (1)
- EU Timber Regulation (1)
- Entscheidung (1)
- Exercise Physiology (1)
- Explainable AI (1)
- Export Credit Agency (1)
- Footwear (1)
- Führung (1)
- Führungskultur (1)
- Gebäudetechnik (1)
- Genossenschaft (1)
- Genossenschaften und genossenschaftliche Rechtsform (1)
- Genossenschaftliche Innovationsökosysteme (1)
- Globalisierung (1)
- GoogleNet (1)
- Haustechnik (1)
- Image Processing (1)
- Industrie (1)
- Innovationssysteme (1)
- KMU (1)
- Klein- und Mittelbetrieb (1)
- Künstliche Intelligenz (1)
- Leitbild (1)
- Locomotion (1)
- Mittelstands-Hubs (1)
- MobileNet (1)
- Nachhaltigkeit (1)
- Organisationskultur (1)
- PM-Entwicklung (1)
- PM-Trends (1)
- Pulmonary (1)
- Pädagogik vs. Lernmaschinen (1)
- ResNet (1)
- Running performance (1)
- Schulentwicklung (1)
- Schulkultur (1)
- Schulprofil (1)
- Spikes (1)
- Sport Science (1)
- Sports performance (1)
- Time Series Classification (1)
- Trade (1)
- Tuberculosis (1)
- Visual Analytics (1)
- Website-Marketing (1)
- Working Capital; Analytics (1)
- Xray (1)
- adversarial attacks (1)
- athletic performance (1)
- biochar (1)
- context-driven design (1)
- cushioning (1)
- energy system analysis (1)
- energy systems modelling (1)
- image classification (1)
- maceration, vessel elements (1)
- maximal sprinting speed (1)
- professional designers (1)
- pyrolysis (1)
- robustness (1)
- socially assistive robot (1)
- super spikes (1)
- track running (1)
- visual qualities (1)
- wood identification (1)
- zero emissions (1)
Institute
- Fakultät Medien (M) (ab 22.04.2021) (13)
- Fakultät Elektrotechnik, Medizintechnik und Informatik (EMI) (ab 04/2019) (10)
- Fakultät Wirtschaft (W) (9)
- IMLA - Institute for Machine Learning and Analytics (9)
- Fakultät Maschinenbau und Verfahrenstechnik (M+V) (5)
- INES - Institut für nachhaltige Energiesysteme (3)
- IfTI - Institute for Trade and Innovation (3)
- IBMS - Institute for Advanced Biomechanics and Motion Studies (ab 16.11.2022) (2)
- ACI - Affective and Cognitive Institute (1)
Open Access
- Open Access (23)
- Closed (15)
- Diamond (12)
- Bronze (10)
- Hybrid (1)
State-of-the-art models for pixel-wise prediction tasks such as image restoration, image segmentation, or disparity estimation, involve several stages of data resampling, in which the resolution of feature maps is first reduced to aggregate information and then sequentially increased to generate a high-resolution output. Several previous works have investigated the effect of artifacts that are invoked during downsampling and diverse cures have been proposed that facilitate to improve prediction stability and even robustness for image classification. However, equally relevant, artifacts that arise during upsampling have been less discussed. This is significantly relevant as upsampling and downsampling approaches face fundamentally different challenges. While during downsampling, aliases and artifacts can be reduced by blurring feature maps, the emergence of fine details is crucial during upsampling. Blurring is therefore not an option and dedicated operations need to be considered. In this work, we are the first to explore the relevance of context during upsampling by employing convolutional upsampling operations with increasing kernel size while keeping the encoder unchanged. We find that increased kernel sizes can in general improve the prediction stability in tasks such as image restoration or image segmentation, while a block that allows for a combination of small-size kernels for fine details and large-size kernels for artifact removal and increased context yields the best results.