Computer Vision (Image classification) using TensorFlow in a simple .NET Core console app.
This sample shows how to use a pre-trained Deep Neural Network TensorFlow called Inception v3 for image classification. This pre-trained model can be downloaded from this URL:
https://storage.googleapis.com/download.tensorflow.org/models/inception5h.zip
This Inception v3 model for image recognition was originally defined in the following paper from the University College London:
http://arxiv.org/abs/1512.00567
The Inception v3 model (which is a Convolutional Neural Network with many layers) has nearly 25 million parameters and uses 5 billion multiply-add operations for classifying a single image. On a modern PC without a GPU this can be done in a fraction of a second per image.