0

squeezenet_demo

Github repo
Main metrics
  • Owner: DT42/squeezenet_demo
  • Platform:
  • License:: GNU General Public License v3.0
  • Category::
  • Topic:
  • Like:
    0
      Compare:

Github stars Tracking Chart

SqueezeNet Keras Implementation

This is the Keras implementation of SqueezeNet using functional API (arXiv 1602.07360).
SqueezeNet is a small model of AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size.
The original model was implemented in caffe.

Reference

pysqueezenet by yhenon

Differences:

  • Switch from Graph model to Keras 1.0 functional API
  • Fix the bug of pooling layer
  • Many thanks to StefOe, the source can now support Keras 2.0 API.

Result

This repository contains only the Keras implementation of the model, for other parameters used, please see the demo script, squeezenet_demo.py in the simdat package.

The training process uses a total of 2,600 images with 1,300 images per class (so, total two classes only).
There are a total 130 images used for validation. After 20 epochs, the model achieves the following:

loss: 0.6563 - acc: 0.7065 - val_loss: 0.6247 - val_acc: 0.8750

Model Visualization

Main metrics

Overview
Name With OwnerDT42/squeezenet_demo
Primary LanguagePython
Program languagePython (Language Count: 1)
Platform
License:GNU General Public License v3.0
所有者活动
Created At2016-06-03 13:52:06
Pushed At2018-05-23 05:29:33
Last Commit At2018-05-23 13:29:32
Release Count0
用户参与
Stargazers Count239
Watchers Count15
Fork Count81
Commits Count26
Has Issues Enabled
Issues Count8
Issue Open Count3
Pull Requests Count5
Pull Requests Open Count0
Pull Requests Close Count0
项目设置
Has Wiki Enabled
Is Archived
Is Fork
Is Locked
Is Mirror
Is Private