September 19, 2024
Last updated
Was this helpful?
Last updated
Was this helpful?
Improved user experience in low connectivity settings
Field engineers can sometimes work in low internet connectivity areas. In this context, reporting times are lengthened, as AI analyses take longer to complete, and sometimes do not go all the way, leading to a poor user experience.
We are releasing a series of improvements to improve the experience in low connectivity settings:
In case the analysis takes too much time to be completed (because of the poor connectivity, the analysis can timeout), the users will be warned that the photo has been stored and that the analysis will be automatically replayed once they have a better connectivity.
We have set the timeout at 30 seconds, meaning that if after 30 seconds the analysis has not been completed therefore we will trigger the flow described in the first bullet.
Version configuration improvements
We've just released a wave of enhancements to the version configuration flow in the Apps Center!
Configuring version is an essential feature to allow a high-level of customization of your application, yet it can be time-consuming. We hope the following enhancements will ease the version configuration process.
App creation simplification: It is not necessary anymore to select the type of application being created.
Automatic redirection to the version page after creating a new version: Instead of being redirected to the application listing after creating a new version, you are now redirected to the version page itself.
Browse images to add example images: Until now, it was only possible to drag and drop example images. It is. not now possible to browse images directly from your device.
Refined usability & design: You may notice a handful of other things that just work a bit better now, with improved consistency across the Apps Center!
Add third-party datasets while launching trainings
In order to increase the robustness of our trained neural networks to without concept images which are out of context of the project, we have developed an ability to add selection of external images in to the training set.
Through this feature, you are able to configure adding a desired number of images from Coco dataset in the training form designated for launching new trainings in the views.
