23 fev 2015 ano - Speeding-up convolutional neural networks using fine-tuned cp-decomposition (ICLR 2014)

Descrição:

Four-component Decomposition



结果：
For the standard ImageNet architecture (AlexNet), the approach speeds up the second convolution layer by a factor of 4x at the cost of 1% increase of the overall top-5 classification error.

Adicionado na linha do tempo:

Data: