vita.aeterna/PetDetectionPy
1
0
Fork 0
mirror of https://github.com/VitaAetaerna/PetDetectionPy.git synced 2025-01-18 07:51:36 +01:00
Code Issues Packages Projects Releases Wiki Activity

Update README.txt

Browse Source
This commit is contained in:
Vita Aeterna 2023-01-05 00:20:38 +01:00 committed by GitHub
parent b7610f4b9e
commit b1b755711a
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 22 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
README.txt
View File

@ -1,19 +1,34 @@
Made by Vita_Aeterna and Google researches.
!!!! Made by Vita_Aeterna and Google researches. !!!!
©Used DataSet "https://www.microsoft.com/en-us/download/details.aspx?id=54765"
Simple tensorflow Neural Network with Sigmoid and with Binary_Crossentropy, learning with and about accuracy and loss.
Simple tensorflow Neural Network with Sigmoid and Binary_Crossentropy, learning with and about accuracy and loss.
All of the Pictures are converted to (50,50) sized Pictures and also GrayScaled (To improve the learning process of the Neural Network).
At first i tested multiple Versions of The Neural Network, The one with the best Accuracy and smallest Loss is used to predict the Animals.
At first i tested multiple Versions(different Dense layers, Conv layers and Layer size) of The Neural Network, The one with the best Accuracy and smallest Loss is used to predict the Animals.
I noticed that if u give an invalid picture like a Man or a Woman the Neural Network will always give the Ouput 0!
As Input a Picture of a Dog or a Cat is given (Need to change Name of your Picture in "EndProduct" script, default Name is "dog.jpg")
Note that the Picture has to be in png or jpg Format.
Path of Cat and Dog Pictures (DataSet) in "ImageUsing" need to be changed too! (DATA_DIR)
As Output, 0 or 1 is given.
0=Dog;
1=Cat;
!0=Dog; 1=Cat!
All packages (pip) used in this Project:
tensorflow / tensorflow-gpu
pickle
numpy
random
os
matplotlib
tqdm
cv2
time