资源分类
》 资源 》算法》caffe-squeezeDet

caffe-squeezeDet

2020-01-14 | |  30 |   0 |   0

caffe-squeezeDet

This is the caffe version of squeezeDet. And I converted tensorflow  model directly into caffemodel.

Note

The convolution operation is different in tensorflow and caffe, especially the padding iterm. Thus, using directly converted model will cause problems. Trick:  using a pad = 2 convolution operation for CONV1 Layer to get a larger feature map and than crop into the right size.

But for kernel size 3 x 3 or 1 x 1 and stride step 1 with padding SAME in tensorflow, we directly choose convolution param with pad = 1, kernel size = 3 and stride = 1

Demo

cd caffe-squeezeDet
python ./src/demo.py


上一篇：squeezedet-keras

下一篇：SqueezeDetTRT

用户评价
全部评价

热门资源

  • Keras-ResNeXt

    Keras ResNeXt Implementation of ResNeXt models...

  • seetafaceJNI

    项目介绍 基于中科院seetaface2进行封装的JAVA...

  • spark-corenlp

    This package wraps Stanford CoreNLP annotators ...

  • capsnet-with-caps...

    CapsNet with capsule-wise convolution Project ...

  • inferno-boilerplate

    This is a very basic boilerplate example for pe...