A Simple Flower classification package trained on DenseNet201
Project description
Keras Flower
A Simple Flower classification package using DenseNet201.
- Classes : 104 classes of flowers
- Training Data Set : Petals to the Metal - Flower Classification on TPU
Usage
- Use pip install to install this package
pip install keras_flower
- To get all prediction results
import keras_flower as kf
predictions = kf.predict_by_path("file/to/predict.png")
print(predictions)
Sample output:
[1.4414026e-06 1.6031330e-06 1.6295390e-06 1.1156463e-06 2.7592062e-06
...
1.1587109e-06 4.1556059e-06 1.0784672e-05 6.0254356e-06]
- To get top prediction result with flower labels
import keras_flower as kf
for predicted, score in kf.predict_name_by_path("/path/to/file.png"):
print(predicted, score)
Sample output:
sunflower 0.99960905
Validation
- Data set : 102 Category Flower Dataset
- No of images : 8189
- Overall accuracy : 0.9715
- confusion matrix : 102flowers_confusion_matrix.csv
- classification report : 102flowers_classification_report.txt
- classification report summary:
precision recall f1-score support
micro avg 0.97 0.97 0.97 8189
macro avg 0.95 0.94 0.94 8189
weighted avg 0.97 0.97 0.97 8189
Demo
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
keras_flower-0.1.tar.gz
(2.6 kB
view hashes)
Built Distribution
Close
Hashes for keras_flower-0.1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 82f64e6e3811658b9cd2fc366df34c7fed6a8b166ba982618bf23bcc9729b7c1 |
|
MD5 | 274a058dd48f8deed47182d539707fab |
|
BLAKE2b-256 | af3ecd8151017259844d4b6ad2db5cc6a3411b7563eb9ad2cd263803f2299dc5 |