search

rupakkarki27 / object_detection

Object Detection App using Flutter and TensorFlow Lite.
58 stars 49 forks source link

Teachable Machine model not working on realtime #2

Open LokieVikky opened 3 years ago

LokieVikky commented 3 years ago

This is the stack trace when i tried to open realtime screen

Caused by: java.lang.IllegalArgumentException: Cannot copy from a TensorFlowLite tensor (Identity) with shape [1, 1] to a Java object with shape [1, 1, 4]. E/AndroidRuntime( 5115): at org.tensorflow.lite.Tensor.throwIfDstShapeIsIncompatible(Tensor.java:482) E/AndroidRuntime( 5115): at org.tensorflow.lite.Tensor.copyTo(Tensor.java:252) E/AndroidRuntime( 5115): at org.tensorflow.lite.NativeInterpreterWrapper.run(NativeInterpreterWrapper.java:175) E/AndroidRuntime( 5115): at org.tensorflow.lite.Interpreter.runForMultipleInputsOutputs(Interpreter.java:360) E/AndroidRuntime( 5115): at sq.flutter.tflite.TflitePlugin$RunSSDMobileNet.runTflite(TflitePlugin.java:651) E/AndroidRuntime( 5115): at sq.flutter.tflite.TflitePlugin$TfliteTask.doInBackground(TflitePlugin.java:448) E/AndroidRuntime( 5115): at sq.flutter.tflite.TflitePlugin$TfliteTask.doInBackground(TflitePlugin.java:422) E/AndroidRuntime( 5115): at android.os.AsyncTask$3.call(AsyncTask.java:394) E/AndroidRuntime( 5115): at java.util.concurrent.FutureTask.run(FutureTask.java:266) E/AndroidRuntime( 5115): ... 4 more I/chatty ( 5115): uid=10509(com.example.object_detection) FinalizerDaemon identical 1 line W/System ( 5115): A resource failed to call destroy. I/Process ( 5115): Sending signal. PID: 5115 SIG: 9

I tried floating point model from teachable machine and still the same issue persists. I even tried my custom model in your sample project and got the same error Im using Tflite.detectObjectOnFrame for detection

flomingdev commented 2 years ago

same problem facing. it detects a black screen and responds to my navigation function and it leads to the crash of the app. if you found any workaround solution then please comment it out. Thank you.