Transformers in Computer Vision

 

[1] Review https://sh-tsang.medium.com/review-vision-transformer-vit-406568603de0

[2] Colab ViT Tutorial https://colab.research.google.com/github/hirotomusiker/schwert_colab_data_storage/blob/master/notebook/Vision_Transformer_Tutorial.ipynb

[3] ViT pytorch https://github.com/lucidrains/vit-pytorch

[4] Vit Keras https://wandb.ai/ayush-thakur/keras_cv_vit/reports/Image-Classification-Using-Vision-Transformer-and-KerasCV--VmlldzozNTE4MzMz

[5] Fine tunning https://huggingface.co/blog/fine-tune-vit

[6] Full material https://github.com/cmhungsteve/Awesome-Transformer-Attention

Watercolor tips

 

Material

[1] https://www.nicoletalle.com/como-elegir-el-mejor-set-de-acuarelas/

Paper

[1] ...

Latex - Format listing environment

minted package 

References:

[1] https://www.overleaf.com/learn/latex/Code_Highlighting_with_minted

Ubuntu 18 OpenGL compilation issues

1) 

In file included from main.cpp:8:0:
gluvi.h:10:10: fatal error: GL/glut.h: No such file or directory
#include <GL/glut.h> // ...when everyone else puts it here?
         ^~~~~~~~~~~

$ sudo apt-get install freeglut3-dev

2)

util.h: In function ‘double randhashd(unsigned int)’:
util.h:259:33: error: ‘UINT_MAX’ was not declared in this scope
{ return randhash(seed)/(double)UINT_MAX; }

#include <climits>

3)


array1.h: In constructor ‘Array1<T>::Array1(long unsigned int, const T*, long unsigned int)’:
array1.h:124:12: error: ‘memcpy’ is not a member of ‘std’
      std::memcpy(data, data_, n*sizeof(T));

#include <cstring>

4) Does not find/recognize libGL

fincahuanaco@pisces:~/DRIVE/Research23/viscosity$ g++ testgl.cpp -lglut -lGLU -L/usr/lib -lGL
/usr/bin/ld: cannot find -lGL
collect2: error: ld returned 1 exit status

#Check symbolic link

$ file /usr/lib/x86_64-linux-gnu/libGL.so 

/usr/lib/x86_64-linux-gnu/libGL.so: broken symbolic link to libGL.so.1.0.0 

 

$ file /usr/lib/x86_64-linux-gnu/libGL.so.1.7.0 #this is right file

/usr/lib/x86_64-linux-gnu/libGL.so.1.7.0: ELF 64-bit LSB shared object, x86-64, version 1 (SYSV), dynamically linked, stripped

#Update libGL link, first go to directory and update

$ cd /usr/lib/x86_64-linux-gnu
$ sudo ln -s libGL.so.1.7.0 libGL.so

#back to source directory and try again

fincahuanaco@pisces:~/DRIVE/Research23/viscosity$ g++ testgl.cpp -lglut -lGLU -lGL            
fincahuanaco@pisces:~/DRIVE/Research23/viscosity$ ./a.out

References:

[1] First OpenGL code http://www.codebind.com/linux-tutorials/install-opengl-ubuntu-linux/

Ubuntu 18: Dark theme

 

$gsettings set org.gnome.desktop.interface gtk-theme 'Adwaita-dark'

References:
[1] https://askubuntu.com/questions/1047604/ubuntu-18-04-missing-global-dark-theme-option-in-gnome-tweak-tool

 

My Wallpapers

 [1] Leopard https://wallhere.com/en/wallpaper/1058505

Graph Neural Network - Step 1

 

References:

[1] Short review https://neptune.ai/blog/graph-neural-network-and-some-of-gnn-applications

[2] Coding GNN https://keras.io/examples/graph/gnn_citations/

[2.1] https://keras.io/examples/graph/gat_node_classification/

AI tools to create videos

[0] Create a Story using chatGPT

[1] Text to speach https://www.speachy.ai  #no includes english/doesnt work

[2] Text and Speach to Video https://invideo.io/

[3] Free Atatars https://www.synthesia.io

Data Visualization tools

 

[1] A new library for LLMs and NNs that provides visual clustering analysis and model interpretability. Super useful.

!pip install arize-phoenix

CG Visualization tools

 

[1] tools https://graphics.tudelft.nl/software/

[2] Surface from fluid in GPU https://github.com/xiaoxiaoyu1872/PostMPS

Latent space

 

References:

[1] https://www.europeanvalley.es/noticias/el-espacio-latente-en-la-ia/

[2] https://rstudio-pubs-static.s3.amazonaws.com/840251_35002e1971484924bc78c24f442e530a.html

Convert Photo in Cartoon

 

 

[1] https://imglarger.com/Cartoonizer

[2] https://vanceai.com/toongineer-cartoonizer/

[3] Prompt image to cartoon https://deepai.org/machine-learning-model/image-editor

[4] https://www.befunky.com/create/photo-to-cartoon/

[5] Head+ [Body models]  https://imagetocartoon.com/ 10/week

[6] 4Adults AI https://metaroids.com/lists/adult-ai-art-tools-that-can-generate-nsfw-ai-images/

WebAssembly

 

[1] Compiling an Existing C Module to WebAssembly

 https://developer.mozilla.org/en-US/docs/WebAssembly/existing_C_to_wasm

Jetson Wifi issues

#Alternative to wpa_supplicant

Creating connection(important don't remove --):

$ nmcli connection add type wifi con-name "myconname" ifname wlan0 ssid "eduroam" -- wifi-sec.key-mgmt wpa-eap 802-1x.eap ttls
 802-1x.phase2-auth mschapv2 802-1x.identity "USERNAME" 

$ sudo --ask nmcli c up myconname 

 

 

$ ls /etc/NetworkManager/system-connections  #show connections

$ cat /etc/NetworkManager/system-connections/myconname #show file connection 

$ nmcli con show myconname  #show connection detailed

Error 1:

password for '802-1x.identity' not given in 'passwd-file' 

Solution 1:

$ vi mypassfile  #add next line with pass value

802-11-wireless-security.psk:mypassvalue

$ nmcli con up <myconname> passwd-file <mypassfile>



References:

[1] Single connection https://www.linuxfordevices.com/tutorials/ubuntu/connect-wifi-terminal-command-line

[2] https://vessokolev.blogspot.com/2018/08/using-nmcli-to-create-8021x-eap-ttls.html

 

 

 

Python - Creating virtual environments

Install pip if it doesn't exist

$curl -O https://bootstrap.pypa.io/get-pip.py

$python3 get-pip.py --user

$python3 -m pip list

$python3 -m pip install tensorflow --no-cache-dir  #for space issue
 

Error during execution of: python3.8 -m venv myenv

Error: Command '['/path/to/desired/env/bin/python3.8', '-Im', 'ensurepip', '--upgrade', '--default-pip']' returned non-zero exit status 1.

Solution: Create a virtual environment using a especific python version

$virtualenv --python=$(which python3.7) envtf4p37

 

Proxy list and tools

 

 

Tools:

[1] https://hidemy.name/en/proxy-checker

[2] Listing proxies by country https://www.proxynova.com/proxy-server-list/country-us

FAPESP Brazil

References

[1] https://fapesp.br/valores/bolsasnopais

Jetson Nano and Tensorflow 2.4 in Python3.6

 

References:

[1] https://qengineering.eu/install-tensorflow-2.4.0-on-jetson-nano.html

[2] Lite https://qengineering.eu/install-tensorflow-2-lite-on-jetson-nano.html

[3] Install bazel https://github.com/jkjung-avt/jetson_nano/blob/master/install_bazel-3.1.0.sh

Linux verify disk activities

 

 

$ sudo apt-get install iotop

$ sudo iotop -a
$ sudo iotop -oPa -d 2  #more specific
 

$ iostat  #activity by sda, sdb, etc.
 

jbd2 is a kernel thread that updates the filesystem journal.

 

References:

[1] Reduce firefox disk access https://gist.github.com/gagarine/5cf8f861abe0dd035b7af19e4f691cd8

[2] Monitoring disk activity(htop) https://askubuntu.com/questions/276669/how-to-monitor-disk-activity

Get create/birth date from files

$ stat wakeup.txt  #get the inode

$ df -h wakeup.txt #get device name sdxy

$ sudo debugfs -R 'stat <586205>' /dev/sda1

Output:

Inode: 586205   Type: regular    Mode:  0664   Flags: 0x80000
Generation: 1277687204    Version: 0x00000000:00000001
User:  1000   Group:  1000   Size: 31
File ACL: 0    Directory ACL: 0
Links: 1   Blockcount: 8
Fragment:  Address: 0    Number: 0    Size: 0
ctime: 0x615eed87:dba39c9c  -- Thu Oct  7 09:52:23 2021
atime: 0x64231159:334e6b48  -- Tue Mar 28 13:10:01 2023
mtime: 0x615eed87:dba39c9c  -- Thu Oct  7 09:52:23 2021
crtime: 0x615eed87:dba39c9c -- Thu Oct  7 09:52:23 2021
Size of extra inode fields: 32

 

References:

[1] https://www.linuxshelltips.com/find-file-creation-date/

[2] https://www.baeldung.com/linux/get-file-creation-date