design choice
Benchmarking the Attribution Quality of Vision Models Robin Hesse 1 Simone Schaub-Meyer 1,2 Stefan Roth 1,2 1 Department of Computer Science, Technical University of Darmstadt
Attribution maps are one of the most established tools to explain the functioning of computer vision models. They assign importance scores to input features, indicating how relevant each feature is for the prediction of a deep neural network.
Country:
- Europe > Germany > Hesse > Darmstadt Region > Darmstadt (0.40)
- Europe > Italy > Marche > Ancona Province > Ancona (0.04)
- Asia > Middle East > Jordan (0.04)
Technology:
ADGym: Design Choices for Deep Anomaly Detection
Many WSAD methods attribute their improvements to novel network architectures or loss functions, based on the authors' understanding of
Country:
- North America > United States > Pennsylvania > Allegheny County > Pittsburgh (0.04)
- Asia > Japan > Honshū > Chūbu > Toyama Prefecture > Toyama (0.04)
- Asia > China > Shanghai > Shanghai (0.04)
Genre:
- Research Report > New Finding (1.00)
- Research Report > Experimental Study (0.93)
Technology:
Country:
- North America > United States > Pennsylvania > Allegheny County > Pittsburgh (0.04)
- Asia > Japan > Honshū > Chūbu > Toyama Prefecture > Toyama (0.04)
- Asia > China > Shanghai > Shanghai (0.04)
Genre:
- Research Report > New Finding (1.00)
- Research Report > Experimental Study (0.93)
Technology:
AUnifiedAnalysisofMixedSampleData Augmentation: ALossFunctionPerspective
Using the theoretical results, we provide a high-level understanding of howdifferentdesign choices ofMSDAworkdifferently.
Country:
- Europe > Netherlands > South Holland > Leiden (0.04)
- Europe > Austria > Tyrol > Innsbruck (0.04)
- Europe > United Kingdom > England > Cambridgeshire > Cambridge (0.04)
Technology:
Table 1: Effect of sharedvs. unshared parameters in the jointencoderf (;θ)oftheTAB-VCRmodel
ResNet101: In Tab. 3 we study the effect of finetuning the last conv block of ResNet101 and the downsample net.31
Technology: Information Technology > Artificial Intelligence > Machine Learning > Neural Networks (0.50)
