Ubuntu instant-off option

Objective: (safe) power-off  the computer with pressing the power button

Refer to:
https://askubuntu.com/questions/430776/how-can-i-modify-the-power-button-behaviour
https://askubuntu.com/questions/1000393/how-to-configure-the-power-off-button-to-just-power-off-instantly-in-ubuntu-17-1?rq=1
https://ubuntuforums.org/showthread.php?t=2270015

3D scalar field plot using MATLAB

refer to: http://matlabnewbie.blogspot.com/2013/07/recently-i-read-post-from-dr.html

Switch gcc versions between 4.9 and 5.4 or over

ref: https://codeyarns.com/2015/02/26/how-to-switch-gcc-version-using-update-alternatives/

after setting
$ sudo update-alternatives --config gcc

How to use another gpu when the first gpu is out of memory

tip: https://github.com/BVLC/caffe/issues/440
http://kawahara.ca/caffe-how-to-specify-which-gpu-to-use-in-pycaffe/

before initializing gpu, set the gpu_id first

Errors in Caffe

Traceback (most recent call last):
  File "Scripts/compute_bn_statistics.py", line 15, in <module>
    import caffe
  File "/home/morinjwang/DL/caffe-segnet-cudnn5/python/caffe/__init__.py", line 1, in <module>
    from .pycaffe import Net, SGDSolver, NesterovSolver, AdaGradSolver, RMSPropSolver, AdaDeltaSolver, AdamSolver
  File "/home/morinjwang/DL/caffe-segnet-cudnn5/python/caffe/pycaffe.py", line 13, in <module>
    from ._caffe import Net, SGDSolver, NesterovSolver, AdaGradSolver, \
ImportError: dynamic module does not define init function (init_caffe)

check if you compile your Caffe library using python 2.7 OR python 3.*

Errors when installing Caffe

설치 하라는거 설치 잘 하고, 위치 정확히 지정해 주고, 에러메시지 잘 해독하면 된다

Tutorial of training ENet using caffe
(https://github.com/TimoSaemann/ENet/tree/master/Tutorial)

+ how to deal with the number of labels in images using cityscapes dataset

No multiple gpu support
(https://github.com/TimoSaemann/ENet/issues/53)

NFS(Network File System)

:without using direct connection via USB cable, or using ssh or minicom

LINK: https://www.fun25.co.kr/blog/linux-ubuntu-16-04-nfs-setting/

Getting 403 forbidden error in sudo apt-get update

sol: resetting apt repository

(LINK: https://ksy3241blog.wordpress.com/2016/06/17/sudo-apt-get-update-%EC%97%90%EB%9F%AC/)

Trial and error to follow SegNet-Tutorial(3) 
(link: http://mi.eng.cam.ac.uk/projects/segnet/tutorial.html)

5. training SegNet using Cityscapes dataset
number of labels in Cityscapes dataset: 34 (id 0 to 33)
re-grouping labels into 11+1 classes ( sky, building, pole, road, pavement, tree, signsymbol, fence, car, pedestrian, bicyclist, unlabelled )

(to-be-updated)

Trial and error to follow SegNet-Tutorial(2)
(link: http://mi.eng.cam.ac.uk/projects/segnet/tutorial.html)

4. test my own data
1) change the path of dense_image_data_param-source in *.prototxt file

========================================================================
segnet_inference_kitti.prototxt
name: "VGG_ILSVRC_16_layer"
layer {
  name: "data"
  type: "DenseImageData"
  top: "data"
  top: "label"
  dense_image_data_param {
    source: "/home/morinjwang/data/JWang2017IROS/test.txt" # Change this to the absolute path to your data file
    batch_size: 1
  }
}
========================================================================

Note that if you want to automatically resize the testing images, try to add two lines below

========================================================================
segnet_inference_kitti.prototxt
name: "VGG_ILSVRC_16_layer"
layer {
  name: "data"
  type: "DenseImageData"
  top: "data"
  top: "label"
  dense_image_data_param {
    source: "/home/morinjwang/data/JWang2017IROS/test.txt" # Change this to the absolute path to your data file
    batch_size: 1
    new_height: 360
    new_width: 480
  }
}
========================================================================

2) make the test.txt file (/path/of/the/original/image /path/of/the/corresponding/groundtruth/image)

========================================================================
text.txt
...
/home/morinjwang/data/JWang2017IROS/image_02/0000/000110.png  /home/morinjwang/data/JWang2017IROS/img_gth/segments/0000/000110.png
...
========================================================================

3) run test_segmentation_camvid.py

Trial and error to follow SegNet-Tutorial(1)
(link: http://mi.eng.cam.ac.uk/projects/segnet/tutorial.html)

1. download caffe-segnet or caffe-segnet-cudnn accelerated version
- choose one considering your ubuntu configuration. In my case, I tried both, however, I used the 'caffe-segnet-cudnn'

(from https://github.com/ethereon/caffe-tensorflow/issues/118)
Note. we SHOULD download above caffe without using pre-installed caffe. because some parameters are deprecated and no longer valid in our current Caffe version (updated on 1.8.2018).
you can see the definitions in ${CAFFE_ROOT}/src/caffe/proto/caffe.proto

Ex) 
message LayerParameter{
  ...
  optional BNParameter bn_param = 45;
  ...
}

message BNParameter{
  ...
  optional FillerParameter scale_filler = 1;  // The filler for the scale
  optional FillerParameter shift_filler = 2;  // The filler for the shift
}
>> those parameters are no longer valid in current version of Caffe.

2. download trained weights
(link: https://github.com/alexgkendall/SegNet-Tutorial/blob/master/Example_Models/segnet_model_zoo.md)
- For testing segmentation performance of urban scenes, download two files:

CAMVID-SegnetBasic:  
http://mi.eng.cam.ac.uk/~agk34/resources/SegNet/segnet_basic_camvid.caffemodel ( segnet_basic_camvid.prototxt)
Cityscapes: 
http://mi.eng.cam.ac.uk/~agk34/resources/SegNet/segnet_iter_30000_timo.caffemodel ( segnet_model_driving_webdemo.prototxt)

3. testing the SegNet
using downloaded weights, we can test the network.
1) run test_segmentation_camvid.py
> python SegNet-Tutorial/Scripts/test_segmentation_camvid.py --model SegNet-Tutorial/Models/segnet_inference.prototxt --weights SegNet-Tutorial/Models/Inference/segnet_iter_30000_timo.caffemodel --iter 233

Note. for pretrained weights using cityscapes, author used 11-class model.

In 'test_segmentation_camvid.py' ignore 4th class 'road_marking'

you need to modify the example script a little bit

========================================================================
Activate conda environment

https://conda.io/docs/user-guide/tasks/manage-python.html

========================================================================
Caffe-segnet-cudnn build error
when uncommenting 'use_cudnn = 1', there is an error as follows:
...
bc: symbol lookup error: /opt/anaconda3/lib/libreadline.so.6: undefined symbol: PC

...

try to install readline properly in conda (with ncurses)
https://github.com/conda-forge/rpy2-feedstock/issues/1

CAFFE + ANACONDA + PYTHON3.* + CUDA8 + OpenCV3 install reference

https://guozhilingblog.wordpress.com/2016/10/10/caffepython3-5anaconda3opencv3-1cuda7-5ubuntu14-04/
https://yangcha.github.io/Caffe-Conda/
http://blog.csdn.net/lien0906/article/details/51784191
http://yingshu.ink/2017/01/12/Python3-5-Anaconda3-Caffe%E6%B7%B1%E5%BA%A6%E5%AD%A6%E4%B9%A0%E6%A1%86%E6%9E%B6%E6%90%AD%E5%BB%BA/
https://hub.docker.com/r/yangcha/caffe-gpu-conda/~/dockerfile/

VIM cheatsheet

(from http://itisfun.tistory.com/281)

Kill processes in ubuntu

ps -ef | grep PROCESSNAME
sudo kill -9 PID

http://121202.tistory.com/45

Instruction of visdom (official github site)

https://github.com/facebookresearch/visdom

Visdom running error (Python3.6.1+anaconda+multiuser env)

Error like "Could not open static file"

1. Try to change the lines in /opt/anaconda3/lib/python3.6/site-packages/visdom/static/index.html

following the link

http://blog.csdn.net/xiaocainiaodeboke/article/details/78941243

(https://github.com/facebookresearch/visdom/issues/185, help)

2. Manually download the missing files. For my case, 'fonts/classnames' and 'fonts/layout_bin_packer' are missing. Try to download the files in the following links

'fonts/classnames': http://jedwatson.github.io/classnames

'fonts/layout_bin_packer': https://github.com/stefanpenner/layout-bin-packer

(Updated 21:29 18.1.6)
because of classnames and layout_bin_packer directories which I made by hand (stupid)

Note that after "pip uninstall visdom", files are not clearly removed in case of the downloaded scripts when you first run visdom.

SSD(Single shot detector) PyTorch implementation github link, tutorial

https://github.com/amdegroot/ssd.pytorch#download-voc2007-trainval--test

GPU usage check in Ubuntu

http://dongjinlee.tistory.com/entry/Check-the-gpu-usage-in-Ubuntu

========================================================================

CUDA version check

nvcc --

========================================================================

CUDNN version check

cat /usr/local/cuda/include/cudnn.h | grep CUDNN_MAJOR -A 2

Permission denied 13 error for running visdom.server

chmod 775 -R /directory/you/want/to/modify

https://askubuntu.com/questions/536392/error-13-permission-denied

Python Numpy sum error (single-shot multibox object detector in PyTorch)

https://stackoverflow.com/questions/46420007/numpy-sum-got-an-keepdims-error

By updating PyTorch version to 0.2 > problem is solved

https://discuss.pytorch.org/t/updating-pytorch/309