tensorflow / tfjs

A WebGL accelerated JavaScript library for training and deploying ML models.
https://js.tensorflow.org
Apache License 2.0
18.44k stars 1.92k forks source link

react-native/pose-detection example is not working #7952

Open rutledmj opened 1 year ago

rutledmj commented 1 year ago

I have been trying to get this example to work:

react-native/pose-detection

But the handleCamerStream here never fires:


<TensorCamera
          ref={cameraRef}
          style={styles.camera}
          autorender={AUTO_RENDER}
          type={cameraType}
          // tensor related props
          resizeWidth={getOutputTensorWidth()}
          resizeHeight={getOutputTensorHeight()}
          resizeDepth={3}
          rotation={getTextureRotationAngleInDegrees()}
          onReady={handleCameraStream}
        />
gaikwadrahul8 commented 1 year ago

Hi, @rutledmj

Thank you for bringing this issue to our attention and I tried to replicate the same issue from my end, and tfjs-examples/react-native/pose-detection tfjs-example yarn and yarn startcommands are working as expected and it's showing the QR code to scan with the help of Expo Goapplication but it's throwing below error message for me

Uncaught Error:java.net.ConnectException: Failed to connect to /127.0.0.1:19000 so May I know are you also getting the same error message which I mentioned above or in your case it's working fine but it's showing output with low FPS ?

If you're getting different error message then please share error output to investigate this issue further. Thank you!

rutledmj commented 1 year ago

Hi @gaikwadrahul8

When I scan the QR code the app launches and hangs. Expo Go says it's taking longer than expected and nothing else occurs.

When I try to run the code via Expo Snack, the app launches and the camera works but handleCameraStream is never invoked.

gaikwadrahul8 commented 1 year ago

Hi, @rutledmj

Thank you for your findings and observations which will help us to investigate this issue further and resolve soon, we will update you soon. Thank you!