py-caffe-cvprw15

Brief Intro

This is a Python reimplementation of this repo. It includes:

• Replacing MATLAB code with PyCaffe during evaluation phase

• Gives exactly the same mAP value for the default caffemodel (KevinNet_CIFAR10_48.caffemodel) on default dataset (cifar-10), i.e. 0.897373

Dependencies

• Caffe with its Python wrapper
• OpenCV's Python wrapper
• Common Python packages(numpy, matplotlib, etc)

Evaluation

1. Prepare dataset and pretrained model as claimed in caffe-cvprw15. You can also download the caffemodel file via BaiduYun with password(3doq)

2. run this bash command:

python py/run_cifar10.py

which will compute mAP and precision at different level, also gives the following output:

Misc

• py/ilsvrc2012_mat_mean.pkl is converted from matlab/caffe/ilsvrc_2012_mean.mat of caffe-cvprw15.

• In MatCaffe, blobs are with order (W,H,C,N), and in PyCaffe, it use (N,C,H,W)