Simple Xcode Swift iOS POC project with machine learning
Budget: €30 – €250 EUR
Goal is to create a simple as possible proof of concept showing the machine learning capabilities on iOS.
You should be able to add sketches like shown in the picture with the items 1 - 19.
In best case these sketches should be added once to the model and if necessary a model should be created from this data.
After this is done a classification should happen for example if I paint something with my finger or a Apple Pencil on PKCanvasView a evaluation should happen.
This view can be easily added like described here:
https://www.raywenderlich.com/12198216-drawing-with-pencilkit-getting-started
Otherwise working just with images is also fine for this POC.
It should be clarified how the source data should look like. Vector? Pixel? Size? Resolution and stuff like that should be requested by you to get best results and best performance for the evaluation.
The result should show me which item was recognised and how well it was recognised.
It also should be always possible to add a invite number of new objects that the project should be able to identify.
You should be able to add sketches like shown in the picture with the items 1 - 19.
In best case these sketches should be added once to the model and if necessary a model should be created from this data.
After this is done a classification should happen for example if I paint something with my finger or a Apple Pencil on PKCanvasView a evaluation should happen.
This view can be easily added like described here:
https://www.raywenderlich.com/12198216-drawing-with-pencilkit-getting-started
Otherwise working just with images is also fine for this POC.
It should be clarified how the source data should look like. Vector? Pixel? Size? Resolution and stuff like that should be requested by you to get best results and best performance for the evaluation.
The result should show me which item was recognised and how well it was recognised.
It also should be always possible to add a invite number of new objects that the project should be able to identify.