Statements (31)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
gptkb:convolutional_neural_network
|
| gptkbp:characteristic |
feature reuse
dense connectivity improved gradient flow reduced parameters |
| gptkbp:describedBy |
gptkb:Densely_Connected_Convolutional_Networks
|
| gptkbp:hasBlockType |
dense block
transition layer |
| gptkbp:hasGrowthRate |
32
|
| gptkbp:hasInputSize |
224x224
|
| gptkbp:hasOutputClasses |
1000 (ImageNet)
|
| gptkbp:hasParameterCount |
20 million (approximate)
|
| https://www.w3.org/2000/01/rdf-schema#label |
DenseNet-201
|
| gptkbp:implementedIn |
gptkb:TensorFlow
gptkb:Keras gptkb:PyTorch |
| gptkbp:introduced |
gptkb:Laurens_van_der_Maaten
gptkb:Gao_Huang gptkb:Kilian_Q._Weinberger gptkb:Zhuang_Liu |
| gptkbp:introducedIn |
2017
|
| gptkbp:level |
201
|
| gptkbp:partOf |
DenseNet family
|
| gptkbp:predecessor |
gptkb:DenseNet-169
|
| gptkbp:pretrainedWeightsAvailable |
gptkb:ImageNet
|
| gptkbp:successor |
gptkb:DenseNet-264
|
| gptkbp:usedFor |
feature extraction
image classification object detection |
| gptkbp:bfsParent |
gptkb:DenseNet
|
| gptkbp:bfsLayer |
6
|