|
gptkbp:instanceOf
|
gptkb:convolutional_neural_network
deep learning model architecture
|
|
gptkbp:arXivID
|
2201.03545
|
|
gptkbp:author
|
gptkb:Zhuang_Liu
gptkb:Trevor_Darrell
gptkb:Saining_Xie
gptkb:Chao-Yuan_Wu
gptkb:Christoph_Feichtenhofer
gptkb:Hanzi_Mao
|
|
gptkbp:bench
|
gptkb:ImageNet
|
|
gptkbp:category
|
gptkb:artificial_intelligence
gptkb:machine_learning
computer vision
|
|
gptkbp:codeAvailableOn
|
gptkb:GitHub
|
|
gptkbp:developedBy
|
gptkb:Facebook_AI_Research
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
ConvNeXt architecture
|
|
gptkbp:improves
|
gptkb:ResNet
|
|
gptkbp:inspiredBy
|
gptkb:Vision_Transformer_(ViT)
|
|
gptkbp:introducedIn
|
2022
|
|
gptkbp:notableFeature
|
layer normalization
GELU activation
increased depth and width
inverted bottleneck blocks
modernized ResNet design
uses large kernel sizes
|
|
gptkbp:notablePublication
|
A ConvNet for the 2020s
|
|
gptkbp:openSource
|
yes
|
|
gptkbp:publishedIn
|
gptkb:arXiv
|
|
gptkbp:usedFor
|
image classification
computer vision tasks
|
|
gptkbp:bfsParent
|
gptkb:ConvNeXt:_Revisiting_ConvNets_for_Image_Recognition
|
|
gptkbp:bfsLayer
|
8
|