Implementation of Hot Dog or not app from Silicon Valley (CNN to identify if the given picture is a hot dog or not)
I used Google's Inception (2015) model. Inception is a deep convolutional neural network built for classifying real world images of thousand category.
Retraining done by replacing last layer of Inception model.
Training has been done using around 300 images so the accuracy for certain images might be low.
- Clone and run bash on this repository. (Ensure that you have TensorFlow installed.)
- Then run
python label_dog.py test/bottle.jpg
- Wait for the model's prediction.
- Here's the result you should expect if things work correctly:
not hot dog : 0.93821
hot dog : 0.06179
- For hot dog run
python label_dog.py test/hotdog.jpg
- The result comes out as
hot dog : 0.99862
not hot dog : 0.00138