Is there a way to give user feedback to the model via the web-bootstrap

[pulse-aj]

I am trying to get more users to click pictures and then inform our model if it's right or wrong in its detection. Is there a way we can use the web-bootstrap to do that?

[mbeissinger]

This is a great idea! It isn't built in to web-bootstrap, but might be worth having in there by default. The question would be where to send the images since everything is in the browser in this starter. Where are you storing/using images?

[mbeissinger]

I would recommend making the predicted label component a dropdown selector (like the 'Use' tab in Lobe) so that users can correct the model by choosing the label directly -- then the image from the canvas element can be sent to a backend server where you store it with the supplied label and can grab into Lobe for more training.

[pulse-aj]

Great! yes, I currently use an S3 bucket and a nodeJS server. Not a coder, but it will be great to have this functionality available as part of this example.

[amyou2005]

anyone success on this task? I looking for this too.