Scene Graph Generation Note 2

本文记录一下Hiwi的相关笔记, 第二部分。

调试工具

PyTorch代码调试利器: 自动print每行代码的Tensor信息

tenserbroad

• for pytorch=1.2
  pip install tensorboard
  pip install tensorflow

• for pytorch=0.4
  pip install tensorboardX

netron

• https://lutzroeder.github.io/netron/

VNC

ssh -fNCT -L 5902:cp202-03:5908 xujing@ssh1.tnt.uni-hannover.de

MSDN

export PYTHONPATH=/home/xujing/Dokumente/MSDN/:/home/xujing/Dokumente/MSDN/faster_rcnn/

export PATH=/home/xujing/miniconda3/envs/msdn/bin:/home/xujing/miniconda3/condabin:/usr/local/cuda-8.0/bin:/home/xujing/.vscode-server/bin/e790b931385d72cf5669fcefc51cdf65990efa5d/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/snap/bin

conda install cudnn=7.1.2

install

Faster RCNN-PyTorch版本运行记录

git checkout pytorch1.0

cd lib
python setup.py build develop

Blog

https://www.telesens.co/2018/03/11/object-detection-and-classification-using-r-cnns/

reference

• 基于深度学习的目标检测技术演进：R-CNN、Fast R-CNN、Faster R-CNN