Maxpooling in SqueezeNet

rcmalli / keras-squeezenet

SqueezeNet implementation with Keras Framework

MIT License

404 stars 148 forks source link

Maxpooling in SqueezeNet #9

Open ChandrahasJR opened 7 years ago

ChandrahasJR commented 7 years ago

The paper seems to have the Maxpooling layers after Fire 4 and Fire 8 modules, but in your implementation they are present after Fire 3 and Fire 8. Any particular reason why this change in architecture ?

mrgloom commented 5 years ago

Any particular reason for using pool_size=(3, 3) in MaxPooling2D and padding='valid' in Conv2D?