☕ a simpler imagenet class synset and decoder utility. pluggable and 100% tf.keras compatible
Project description
☕ amerikana: keras prediction decoder for imagenet-simple-labels
a simpler, more human readable imagenet synset for keras.
comparison:
ID | ImageNet | Keras | Simple (this repo) |
---|---|---|---|
87 | African grey, African gray, Psittacus erithacus | African_grey | grey parrot |
97 | drake | drake | duck |
913 | wreck | wreck | shipwreck |
930 | French loaf | French_loaf | baguette |
quickstart:
should be used as a drop in replacement for tf.keras.applications.imagenet_utils.decode_predictions
install with pip:
pip install amerikana
import:
from amerikana import decode_predictions
instead of:
from tensorflow.keras.applications.imagenet_utils import decode_predictions
example use:
predictions = model.predict(processed_image)
labels = decode_predictions(predictions)
print(labels)
credit:
-
simplified labels list: imagenet-simple-labels. see information and comparison for simplified labels there.
-
the module code used in this project is a derivative of tf.keras imagenet_utils.
enjoy!
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
amerikana-1.1.tar.gz
(3.0 kB
view hashes)