Implementation of Hot Dog or not app from Silicon Valley (CNN to identify if the given picture is a hot dog or not)
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.
python label_dog.py test/bottle.jpg
not hot dog : 0.93821
hot dog : 0.06179
python label_dog.py test/hotdog.jpg
hot dog : 0.99862
not hot dog : 0.00138