|
1.安装anaconda(Python 3.4+)
2.安装tensorflow 1.3+安装tensorflow:对于CPU版本:pip install tensorflow对于GPU版本:pip install tensorflow-gpu升级tensorflow到最新版1.4.0:pip install --upgrade tensorflow-gpu
3.安装keras(2.08+)pip install keras
4.安装scikit-image
sudo pip install scikit-image
5.下载预训练的coco权重文件(mask_rcnn_coco.h5)https://github.com/matterport/Mask_RCNN/releases
6.为了在MS COCO上训练或者测试,需要安装pycocotools.Linux: https://github.com/waleedka/cocoWindows: https://github.com/philferriere/cocoapi.需要安装VC++2015
7.COCO子数据集,5K的minival和35K的validation-minus-minival。minival:https://dl.dropboxusercontent.com/s/o43o90bna78omob/instances_minival2014.json.zip?dl=0validation-minus-minival:https://dl.dropboxusercontent.com/s/s3tw5zcg7395368/instances_valminusminival2014.json.zip?dl=0
|
|