ai.djl.basicmodelzoo.cv.classification

Class SqueezeNet

java.lang.Object

ai.djl.basicmodelzoo.cv.classification.SqueezeNet

public final class SqueezeNet
extends java.lang.Object

SqueezeNet contains a generic implementation of Squeezenet adapted from [torchvision implmentation](https://github.com/pytorch/vision/blob/master/torchvision/models/squeezenet.py)

Squeezenet is an efficient NN used for Image classification. It provides both performance boost and a tiny size. It's a good choice to adopt squeezenet for application runs on Mobile or Edge devices. Implementing the original squeezenet from Forrest N. Iandola, Song Han, Matthew W. Moskewicz, Khalid Ashraf, William J. Dally, Kurt Keutzer. "SQUEEZENET: ALEXNET-LEVEL ACCURACY WITH 50X FEWER PARAMETERS AND 0.5MB MODEL SIZE"

Method Summary

Modifier and Type	Method and Description
static ai.djl.nn.Block	squeezenet(int outSize) Construct squeezenet v1.1.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Detail

squeezenet

public static ai.djl.nn.Block squeezenet(int outSize)

Construct squeezenet v1.1.

Parameters:

outSize - the number of output classes

Returns:

squeezenet Block