image-annotator-docker

Requirements

git clone this repository
image-annotator/ directory to be placed in image-annotator-docker/
- git clone image-annotator/ to image-annotator-docker/ from https://bitbucket.org/JialinYuan/image-annotator/src/master/ (no install needed)
- Download and extract model.zip to image-annotator-docker/image-annotator/deep_interactive/
  - https://oregonstate.box.com/s/dq4tc94f1299lml52a7bh7dztjkhvrj5
Modify image-annotator-docker/image-annotator/config.py
- find line: __C.PosNeg_Model_weight_Path = './deep_interactive/model/PASCAL/pos_neg/model.ckpt-80000' and change it to: __C.PosNeg_Model_weight_Path = '/image-annotator/deep_interactive/model/PASCAL/pos_neg/model.ckpt-80000
cuDNN 7.1.4 and place "cudnn-9.0-linux-x64-v7.1.tgz" in image-annotator-docker/
- https://developer.nvidia.com/rdp/cudnn-archive
- Select "Download cuDNN v7.1.4 (May 16, 2018), for CUDA 9.0"
- Download "cuDNN v7.1.4 Library for Linux"
Docker to be installed on host machine
- https://docs.docker.com/engine/install/ubuntu/
- Complete postinstall steps: https://docs.docker.com/engine/install/linux-postinstall/
nvidia-docker (2.0) installed on host machine (For Ubuntu os, it is debian based distribution)
- https://github.com/nvidia/nvidia-docker/wiki/Installation-(version-2.0)