I want to write to you ,but I don't know your email. So I change the way about flower-recognition HOT 3 CLOSED

1. To prepare the dataset, you need to download FLOWERS17 dataset here. You need to create a folder named flowers and inside that create sub-folders named "daffodil", "snowdrop", "lilyvalley", "bluebell", "crocus", "iris", "tigerlily", "tulip", "fritillary", "sunflower", "daisy", "coltsfoot", "dandelion", "cowslip", "buttercup", "windflower", "pansy"
   Then, you can run the organize.py script. Be aware of input_path and output_path found in that script where you need to specify the dataset path.

2. Imagenet is a very big dataset. Normally, if you don't have GPUs, it is very hard to train on Imagenet. That is why we go for using pre-trained deep learning models for our own dataset.

3. You need to have proper dataset having all the flower-species folder with images inside it. You can then execute extract-features.py script to extract features for a specific pre-trained model.

For more information about this project, you can see my post here.

from flower-recognition.

Thank you for your reply and support , I am going to continue trying.

from flower-recognition.

Hey even I have a doubt about the project. How can I contact you

from flower-recognition.