Hi all,
I kindly expect your help on this. This may be a silly question, since I am new here.
1. I already have an image data set comprised of six classes(separate folders) uploaded to my notebook.
I want to apply all kind of transformations such as
do_flip, flip_vert, max_rotate, max_zoom, max_lighting, max_warp, zoom_crop, rand_resize_crop, brightness , contrast, crop_pad, dihedral, pad etc.
and save in separate folders.
2. When I apply transform during data_bunch creation, I could not include the transforms such as zoom_crop together with the other transforms such as flip_vert . It returns error. I could not understand it. Please help me. Here is my code. tfms = get_transforms( do_flip = True, flip_vert = True, max_rotate = 60.0, max_zoom = 1.1, max_lighting = 0.2, max_warp = 0.2, p_affine = 0.75, p_lighting = 0.75)
Hey @Anuraaj, I am on the way to work so cannot answer your questions fully but I have been working on transformations and saving images to different folders (with respect to results from a confusion matrix).
Hi @amritv I have cloned your Vision_UI found in github and ran the notebook 01_UI_Fastai. But when I click choose folder the following error was returned. Anything should I include further?
@amritv I did install ipywidgets usin pip install (conda install was returning an error and could not install that way) but still the same issue is there.
Hey @Anuraaj - checking to see if you are still getting the error? BTW were you using Colab as I got the same error when I was working on getting vision UI to work on Colab. The issue was with the incompatibility of tkinter and colab.