android / android/codelab-mlkit-android

Object detection and Labeling avoids people classification?

Open
#22 1 comment 0 reactions 0 assignees View on GitHub
Dominant language
Kotlin
Stars
186
Forks
104
PR merge metrics
No merged PRs in 30d

Description

My observation from testing the ML kit is that it seems not to identify category/object Person. Is this a bug, model issue or deliberate filtering? It seems the kit was not aimed to classify object type person, since there are (only :-() 5 object categories defined , at least what I see in the class. BTW why the Object ID integer mapping is not documented in the tutorials? It took me while to identify the category names, since the API returns only integer at the moment

![image](https://user-images.githubusercontent.com/13347592/61459844-eb8e1f00-a96d-11e9-99da-254136a4d99f.png)

Regardless of the fact t( it is rather poor capability for object detection API to be useful), I am not sure why People are committed?

see the details of the issue I have with the ML kit here:

[https://stackoverflow.com/questions/57078901/ml-kit-for-firebase-object-recognition-categories](url)

Contributor guide

Open the contributing guide

Research direction

Start by reviewing the ML Kit object detection behavior described in the issue and the linked Stack Overflow discussion, then inspect the referenced class and tutorial material for the five categories and integer IDs. Determine whether person classification is supported or intentionally filtered, and whether the ID mapping is missing from the tutorials. Done means the behavior and category mapping have a documented, verified explanation.

Written by the indexing model from the issue text.

Assessment

Tech stack
android, kotlin, machine-learning
Domain
documentation, machine-learning, mobile-dev
Issue type
Bug
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
20/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.