SqueezeNet

2000년 도입 · 논문 97편에서 사용

SqueezeNet is a convolutional neural network that employs design strategies to reduce the number of parameters, notably with the use of fire modules that "squeeze" parameters using 1x1 convolutions.

출처: SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size

소개 논문: SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size