Classify Satellites

[ritwik12]

We already have Moon classification. It would be better to have more Satellite.

[alisaffari97]

Can you give me a point to start?

[ritwik12]

@alisaffari97 Thanks for showing interest, You can go through #26 and see what changes were made.
Being an Inception based model you don't need to change Any code in Model just need to provide Training data and some other code changes for more information.
After adding all those if you retrain the model then you can see that it is able to predict new category of data also.
Names of Natural Satellites: https://web.archive.org/web/20070227165513/http://www.planetary.org/explore/topics/compare_the_planets/moon_numbers.html
You can take few names from here

[dejesusj]

Is this still open? If so, I'd like to work on it!

[ritwik12]

Yes this is open @dejesusj :)

[vinodlakshmanan]

Is this still open? I'd like to give it a shot if it is.

[ritwik12]

@vinodlakshmanan Sure, go for it.

[ARGF0RCE]

Hello, I am a newbie in open source but I am experienced with TensorFlow and deep learning in general including data preprocessing. I wish to contribute to this project. May I know how I can help in this issue?

[ritwik12]

@ARGF0RCE For this you won't have to change much in the code. just add new data regarding any new satellites. Currently, we have Moon only. Later train the model using that data and add a new documentation for that.

[Overlrd]

HI @ritwik12 I'm a new contributor and i'm actually collecting images of jupyter's moon Europa and i want ot ask if i have to collect images of a specific shape , or reshaping is already done in preprocessing steps.
I'm Very glad to help

[ritwik12]

@Overlrd Resizing is done. what do you mean by reshaping? You don't need to worry about resolution and dimensions of image. it's take care of.

[Overlrd]

ok it's what i want to know, thanks Le lun. 7 nov. 2022 17:21, Ritwik Sharma ***@***.***> a écrit :

…

@Overlrd <https://github.com/Overlrd> Resizing is done. what do you mean by reshaping? You don't need to worry about resolution and dimensions of image. it's take care of. — Reply to this email directly, view it on GitHub <#39 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AVRSKOHGU6OYGJRWXIXHMILWHE3CDANCNFSM4JKF5KJQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[Overlrd]

I'm getting this error :

File "C:\Users\Sudo Overloord\Overlrd\Celestial_bodies_detection\Celestial-bodies-detection\hub\examples\image_retraining\retrain.py", line 1063, in <module> tf.app.run(main=main, argv=[sys.argv[0]] + unparsed) File "C:\Users\Sudo Overloord\Overlrd\Celestial_bodies_detection\custom_env\lib\site-packages\tensorflow\python\platform\app.py", line 36, in run _run(main=main, argv=argv, flags_parser=_parse_flags_tolerate_undef) File "C:\Users\Sudo Overloord\Overlrd\Celestial_bodies_detection\custom_env\lib\site-packages\absl\app.py", line 308, in run _run_main(main, args) File "C:\Users\Sudo Overloord\Overlrd\Celestial_bodies_detection\custom_env\lib\site-packages\absl\app.py", line 254, in _run_main sys.exit(main(argv)) File "C:\Users\Sudo Overloord\Overlrd\Celestial_bodies_detection\Celestial-bodies-detection\hub\examples\image_retraining\retrain.py", line 785, in main class_count = len(image_lists.keys()) AttributeError: 'NoneType' object has no attribute 'keys'

It seems like the image_list object returned by the def create_image_lists is not a dictionnary
What should i do here ?

[ritwik12]

@Overlrd Please check

  # Look at the folder structure, and create lists of all the images.
  image_lists = create_image_lists(FLAGS.image_dir, FLAGS.testing_percentage,
                                   FLAGS.validation_percentage)
  class_count = len(image_lists.keys())

It is accessing the image folder as passed in the command --image_dir=./training_data, please check if the folder names are correct as they should be. Try to print this image_lists variable and see what does it contains. could be some issue with directory or maybe permission issues.

[Overlrd]

Ok, I'll check. Le sam. 12 nov. 2022 20:10, Ritwik Sharma ***@***.***> a écrit :

…

@Overlrd <https://github.com/Overlrd> Please check # Look at the folder structure, and create lists of all the images. image_lists = create_image_lists(FLAGS.image_dir, FLAGS.testing_percentage, FLAGS.validation_percentage) class_count = len(image_lists.keys()) It is accessing the image folder as passed in the command --image_dir=./training_data, please check if the folder names are correct as they should be. Try to print this image_lists variable and see what does it contains. could be some issue with directory or maybe permission issues. — Reply to this email directly, view it on GitHub <#39 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AVRSKOBEKLBCQSTVTII7H5TWH72T5ANCNFSM4JKF5KJQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>