Statements (26)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
gptkb:convolutional_neural_network
|
| gptkbp:application |
computer vision
image classification |
| gptkbp:architecture |
gptkb:convolutional_neural_network
|
| gptkbp:arXivID |
1605.07146
|
| gptkbp:basedOn |
gptkb:ResNet
|
| gptkbp:bench |
gptkb:SVHN
gptkb:CIFAR-10 gptkb:CIFAR-100 gptkb:ImageNet |
| gptkbp:developedBy |
gptkb:Sergey_Zagoruyko
Nikos Komodakis |
| gptkbp:influenced |
modern CNN architectures
|
| gptkbp:influencedBy |
deep residual learning
|
| gptkbp:mainDifferenceFrom_ResNet |
increased width instead of depth
|
| gptkbp:notableFeature |
faster training
better accuracy with fewer layers |
| gptkbp:notablePublication |
Wide Residual Networks
|
| gptkbp:openSource |
gptkb:TensorFlow
gptkb:PyTorch |
| gptkbp:publishedIn |
2016
|
| gptkbp:url |
https://arxiv.org/abs/1605.07146
|
| gptkbp:uses |
residual connections
|
| gptkbp:bfsParent |
gptkb:TorchVision
|
| gptkbp:bfsLayer |
7
|
| https://www.w3.org/2000/01/rdf-schema#label |
Wide ResNet
|