[PAID] ImageLabelling: Label images with ML Kit [$8 or INR 551]

Extensions
1

1. Introduction

Description: You can use ML Kit to label objects recognized in an image. The default model provided with ML Kit supports 400+ different labels.
Latest Version: 2
Released: 2023-09-23T18:30:00Z
Last Updated: 2023-10-30T18:30:00Z

2. Blocks

image
image448×714 22.7 KB

image

3. Usages

blocks (15)
blocks (15)1683×702 86.3 KB

4. Docs

Events

DepsInstalled Event raised after dependencies has been installed. Image Labeler can be used after this event.
ErrorOccurred Event raised when any error occurs
errorMsg | text
DepsDownloadProgress Event raised when dependency download progress changes
progress | number
LabelingSuccessful Event raised after getting labels from images
labelDict | dictionary
LabelingFailed Event raised if image labelling fails
errorMsg | text

Methods

IsOperational Return whether Image Labeler is operational and can be used
InstallDeps It tells Play Service to install the required dependencies
Initialize Initialize Image Labeler
GetLabels Get image labels from provided file uri
image | any

Properties

ConfidenceThreshold Sets min confidence threshold. Labels below this value will be discarded from final result.
Property Type : write-only
Accepts : number

5. Preview

6. Purchase

PayPal:

Upi:
vknow360@apl

Thank you.
Hope it helps!

6 Likes
2

ChangeLog Version 2

  • Removed DeleteDeps block
  • GetLabels block now accepts Image component, Content Uri and File path
  • Better dependency management
  • Efficiency increased
  • Aix size reduced to 778 kb
  • Offline purchase verification
3

ChangeLog Version 3

  • Added custom model support
    Now you can use a custom tflite model for labelling.
    image

If modelPath starts with / then it will be loaded from storage otherwise from assets.

screenshot-ai2.appinventor.mit.edu-2023.12.26-22_14_01
screenshot-ai2.appinventor.mit.edu-2023.12.26-22_14_011137×220 20.6 KB

  • Added MaxResultsCount property to limit no. of items returned in result.
    Default value is 10

  • Some internal changes

4

Can this detect humans?
I mean if someone takes a photo, i want to check whether people is in it and then upload, else through an error

5

Not with base model, at least I wasn't not able to label persons.

With custom model it definitely can.

1 Like
6

Is this still available for purchase? and can I use this to detect Poker playing card values? thanks

7

You might be able to detect poker cards but this will not read the values.
You need text recognition extension extension for that.