mgyong / awesome-mediapipe Goto Github PK

I want to use mediapipe in react-native, is there any example that I can refer to?
Or is it possible to guide you on how to apply the android example to react-native?

You may want to add these demos to the list of cool and useful MediaPipe demos.

The are using the OSSDC VisionAI realtime video processing platform, see more details here:
https://github.com/OSSDC/OSSDC-VisionAI-Core

More details in each video description,

With processing running on Google Colab free GPU:

Isn't this fun?! MediaPipe Holistic neural net model processed in real time on Google Cloud
https://www.youtube.com/watch?v=0l9Bb5IC86E

With processing running on Myriad X the core of DepthAI cameras

MediaPipe hands tracking on DepthAI camera through OSSDC VisionAI on video stream from phone
https://www.youtube.com/watch?v=CdvpjKfNUWw

MediaPipe hands tracking on DepthAI camera through OSSDC VisionAI
https://www.youtube.com/watch?v=poTrqPBsn4s

More MediaPipe demos will come.

Here are all the OSSDC VisionAI demos:
https://www.youtube.com/watch?v=6a6bqJiZuaM&list=PLUop7b1Q1uZnneN4utdBh7OLYbKJlieQf

Hi there,great job! @mgyong

The blog article on https://ai.googleblog.com/2020/03/real-time-3d-object-detection-on-mobile.html describes a "novel annotation tool for use with AR session data" that was used to create the annotated dataset. Which tool is that? Is that tool also available, so that we can 3D annotate our own AR sessions on custom objects?

I read the description of the dataset preparation tool, but it does not describe 3D annotation.
from here

In the README, Handsfree.js.org redirects to a 302 page. The correct URL is → this.
Moreover, here, it should be TFlite instead of TFllite, & hence this also needs to be fixed.

cc: #16 ; refer to this. // @mgyong

• Yes, I'd love to open a PR to fix this (if approved).