Refine
H-BRS Bibliography
- yes (1877) (remove)
Departments, institutes and facilities
- Fachbereich Informatik (820)
- Fachbereich Angewandte Naturwissenschaften (446)
- Fachbereich Ingenieurwissenschaften und Kommunikation (305)
- Institut für Technik, Ressourcenschonung und Energieeffizienz (TREE) (275)
- Institute of Visual Computing (IVC) (251)
- Fachbereich Wirtschaftswissenschaften (204)
- Institut für funktionale Gen-Analytik (IFGA) (159)
- Institut für Cyber Security & Privacy (ICSP) (83)
- Institut für Verbraucherinformatik (IVI) (79)
- Graduierteninstitut (51)
Document Type
- Conference Object (831)
- Article (629)
- Part of a Book (143)
- Preprint (68)
- Doctoral Thesis (52)
- Book (monograph, edited volume) (41)
- Research Data (22)
- Master's Thesis (22)
- Report (17)
- Conference Proceedings (16)
- Working Paper (15)
- Contribution to a Periodical (6)
- Bachelor Thesis (5)
- Book review (4)
- Lecture (2)
- Study Thesis (2)
- Other (1)
- Patent (1)
Year of publication
Language
- English (1877) (remove)
Has Fulltext
- no (1877) (remove)
Keywords
- Robotics (13)
- FPGA (12)
- Virtual Reality (12)
- Machine Learning (9)
- GC/MS (8)
- ICT (8)
- Quality diversity (8)
- virtual reality (8)
- Performance (7)
- 3D user interface (6)
Dynamic Programming
(2024)
Queueing Theory
(2024)
The Decision Tree Procedure
(2024)
Heuristic Methods
(2024)
Network Analysis Method
(2024)
The Peren-Clement Index
(2024)
Sequencing Problems
(2024)
Linear Optimization
(2024)
The Peren Theorem
(2024)
Pyrolysis–Gas Chromatography
(2024)
The methodology of analytical pyrolysis-GC/MS has been known for several years, but is seldom used in research laboratories and process control in the chemical industry. This is due to the relative difficulty of interpreting the identified pyrolysis products as well as the variety of them. This book contains full identification of several classes of polymers/copolymers and biopolymers that can be very helpful to the user. In addition, the practical applications can encourage analytical chemists and engineers to use the techniques explored in this volume.
Social policy research on the ageing workforce from the perspective of employees and employers
(2024)
In vision tasks, a larger effective receptive field (ERF) is associated with better performance. While attention natively supports global context, convolution requires multiple stacked layers and a hierarchical structure for large context. In this work, we extend Hyena, a convolution-based attention replacement, from causal sequences to the non-causal two-dimensional image space. We scale the Hyena convolution kernels beyond the feature map size up to 191$\times$191 to maximize the ERF while maintaining sub-quadratic complexity in the number of pixels. We integrate our two-dimensional Hyena, HyenaPixel, and bidirectional Hyena into the MetaFormer framework. For image categorization, HyenaPixel and bidirectional Hyena achieve a competitive ImageNet-1k top-1 accuracy of 83.0% and 83.5%, respectively, while outperforming other large-kernel networks. Combining HyenaPixel with attention further increases accuracy to 83.6%. We attribute the success of attention to the lack of spatial bias in later stages and support this finding with bidirectional Hyena.