FoodVision

Click to upload or drag and drop

SVG, PNG, JPG or GIF

ResNet-18

ResNet-18 is a convolutional neural network that is 18 layers deep. It is known for its residual learning framework to ease the training of networks that are substantially deeper than those used previously.

Accuracy
92.5%
Inf. Time
~45ms
Parameters
11.7 M
Model Size
45 MB
Training Set
101,000