I want to build and app for the restaurant that will recognise food on the clients trays. What should I start from? I'm a web developer, so it's quite a new theme for me. First I was thinking of Amazon Recognition or Google Vision but still not sure.
Looks like quite an ambitious project. Generally, you should take a model that recognise not only the food but the objects themselves (like Amazon Recognition, Google Vision, Yolo) and apply transfer learning on the last few layers.
Thanks! I guess this will be my main goal for now, so maybe 6 months full-time. Also, I thought Amazon Recognition uses its own database for object detection. Are you saying I can set my own objects and train Amazon by myself, did I understand you right?
If you have access to the model that you can load on any ML framework, then you can retrain the model with your own image dataset. All you need to do is to change the output layer with the number of the class you need. Look at the transfer learning or fine tuning.
Thanks for your recommendations, I will.
Just drop us an email to ...
Show more 
Log in with Google