Automated search discovers Swish activation f(x) = x * sigmoid(βx) that improves top-1 ImageNet accuracy over ReLU by 0.9% on Mobile NASNet-A and 0.6% on Inception-ResNet-v2.
What is the best multi-stage architecture for object recognition? In 2009 IEEE 12th International Conference on Computer Vision
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.NE 1years
2017 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Searching for Activation Functions
Automated search discovers Swish activation f(x) = x * sigmoid(βx) that improves top-1 ImageNet accuracy over ReLU by 0.9% on Mobile NASNet-A and 0.6% on Inception-ResNet-v2.